Segmented controlled heat welding method and system for dual phase steel thick walled components

Abstract

The application provides a segmented heat control welding method and system for a dual-phase steel thick-wall component, relates to the technical field of welding, and the method comprises the following steps: constructing a segmented unit planning set; generating a segmented unit multi-source synchronous data set; constructing a welding heat cycle characteristic parameter set; generating a welding control constraint set; obtaining a segmented heat input deviation instruction set; generating a welding process parameter adjustment instruction; and according to the actual weld structure performance in a long-term welding process, a parameter optimization algorithm is used to optimize and adjust the process, so that the segmented heat control welding model is continuously updated. The technical problems of extensive heat input control, incomplete temperature field monitoring, process parameter adjustment lag and lack of quantitative basis in the prior art are solved.

Description

Technical Field

[0001] This application relates to the field of welding technology, and in particular to a segmented controlled-heat welding method and system for thick-walled duplex steel components. Background Technology

[0002] Duplex steel, due to its excellent strength, toughness, and corrosion resistance, is widely used in thick-walled load-bearing components in marine engineering, petrochemicals, pressure vessels, and large bridges. Welding, as the core process for manufacturing and connecting these components, directly determines their structural integrity and service safety. Traditional welding methods primarily involve pre-setting constant process parameters and monitoring certain macroscopic parameters during the process to roughly control the welding heat input.

[0003] However, existing technologies have significant limitations for thick-walled duplex steel components with complex geometries and significant thickness variations. Firstly, precise and adaptive control of welding heat input is difficult. Spatial variations in component thickness distribution, bevel angle, and weld trajectory lead to substantial differences in local heat dissipation conditions. Using uniform welding parameters easily results in uneven heat input across different areas, causing problems such as distorted welding stress distribution, uncontrolled interpass temperature, and excessive cooling rate fluctuations, severely impacting the uniformity and mechanical properties of the weld microstructure. Secondly, existing monitoring methods largely rely on single-point or limited-point temperature measurements, failing to comprehensively and synchronously acquire information on the surface and interpass temperature field distribution of the weld pool and surrounding areas. This results in incomplete characterization of the thermal cycle process and difficulty in accurately assessing the deviation between the actual thermal state and the preset process window. Thirdly, adjustments to welding parameters often depend on operator experience or post-processing inspection, lacking a closed-loop feedback control mechanism based on real-time multi-source data fusion. This makes it impossible to dynamically and accurately compensate for heat input deviations during welding, compromising process stability and consistency.

[0004] In summary, existing welding technologies for thick-walled duplex steel components suffer from several technical problems when dealing with non-uniform heat dissipation conditions caused by complex geometries. These problems include crude heat input control, incomplete temperature field monitoring, and delayed and quantitatively unbased adjustments to process parameters, which can easily lead to fluctuations in welding quality and potential defects. Therefore, a method is urgently needed to address these issues. Summary of the Invention

[0005] This disclosure provides a segmented heat-controlled welding method and system for thick-walled duplex steel components, which solves the technical problems in the prior art, such as crude heat input control, incomplete temperature field monitoring, and lagging process parameter adjustment without quantitative basis.

[0006] According to a first aspect of this disclosure, a segmented controlled-heat welding method for thick-walled duplex steel members is provided, comprising: Based on the geometric information of the welding area of ​​the duplex steel thick-walled component, welding segment units are divided, and a segment unit planning set is constructed by combining the process constraints of the segment units. A sensor acquisition system is deployed to collect temperature information and welding process parameters of all welding segment units in real time, and generate a multi-source synchronous dataset of segment units. The sensor acquisition system includes an infrared temperature measurement array and a thermocouple sensor array. The temperature information includes surface temperature and interlayer temperature. The temperature information in the segmented unit multi-source synchronous dataset is preprocessed to extract thermal cycle feature parameters and construct a welding thermal cycle feature parameter set. Based on material properties and welding procedure qualification reports, the planned process parameters, preset line energy and thermal cycle characteristic parameter ranges during the welding process are obtained, and a welding control constraint set is generated. Based on the segmented unit multi-source synchronous dataset, the welding thermal cycle characteristic parameter set and the welding control constraint set, the comprehensive deviation of the thermal input of all welding segmented units is calculated to obtain the segmented thermal input deviation instruction set. Using the segmented thermal input deviation instruction set, the final current adjustment and final speed adjustment are calculated, and welding process parameter adjustment instructions are generated. Based on the actual weld microstructure and properties during long-term welding, a parameter optimization algorithm is used to optimize and adjust the process, thereby achieving continuous updating of the segmented heat-controlled welding model.

[0007] According to a second aspect of this disclosure, a segmented thermally controlled welding system for duplex steel thick-walled members is provided, comprising: The segmented unit planning and construction module is used to divide the welding segmented units according to the geometric information of the welding area of ​​the duplex steel thick-walled component, and construct the segmented unit planning set in combination with the segmented unit process constraints. A multi-source sensor acquisition module is used to deploy a sensor acquisition system to collect temperature information and welding process parameters of all welding segment units in real time, and generate a multi-source synchronous dataset of segment units. The sensor acquisition system includes an infrared temperature measurement array and a thermocouple sensor array. The temperature information includes surface temperature and interlayer temperature. The thermal cycling feature extraction module is used to preprocess the temperature information in the segmented unit multi-source synchronous dataset, extract thermal cycling feature parameters, and construct a welding thermal cycling feature parameter set. A welding process constraint generation module is used to obtain planned process parameters, preset line energy and thermal cycle characteristic parameter ranges during the welding process based on material properties and welding process qualification reports, and generate a welding control constraint set. The segmented thermal input deviation calculation module is used to calculate the comprehensive deviation of the thermal input of all welding segment units based on the segmented unit multi-source synchronous dataset, the welding thermal cycle characteristic parameter set and the welding control constraint set, and to obtain the segmented thermal input deviation instruction set. A process parameter adjustment instruction generation module is used to calculate the final current adjustment amount and the final speed adjustment amount using a segmented heat input deviation instruction set, and generate welding process parameter adjustment instructions. The model optimization and update module is used to optimize and adjust the process based on the actual weld microstructure and properties during long-term welding, thereby achieving continuous updating of the segmented heat-controlled welding model.

[0008] One or more technical solutions provided in this disclosure have at least the following technical effects or advantages: Based on the geometric information of the welding area of ​​the duplex steel thick-walled component, welding segment units are divided; combined with the process constraints of the segment units, a segment unit planning set is constructed; a sensor acquisition system is deployed to collect temperature information and welding process parameters of all welding segment units in real time, generating a segment unit multi-source synchronous dataset; the sensor acquisition system includes an infrared temperature measurement array and a thermocouple sensor array; the temperature information includes surface temperature and interpass temperature; the temperature information in the segment unit multi-source synchronous dataset is preprocessed to extract thermal cycle characteristic parameters, and a welding thermal cycle characteristic parameter is constructed. This method involves several steps: First, based on material properties and welding procedure qualification reports, planned process parameters, preset heat input, and thermal cycle characteristic parameter ranges are obtained during the welding process to generate a welding control constraint set. Second, based on the multi-source synchronous dataset of segmented units, the welding thermal cycle characteristic parameter set, and the welding control constraint set, the comprehensive deviation of the heat input for all welding segmented units is calculated to obtain a segmented heat input deviation instruction set. Third, using this instruction set, the final current adjustment and final speed adjustment are calculated to generate welding process parameter adjustment instructions. Finally, based on the actual weld microstructure and properties during long-term welding, a parameter optimization algorithm is used to optimize the adjustment process, enabling continuous updating of the segmented heat-controlled welding model. This method solves the technical problems of coarse heat input control, incomplete temperature field monitoring, and lagging process parameter adjustment lacking quantitative basis in existing technologies. It achieves refined, adaptive closed-loop control of the welding thermal process, effectively improving the microstructure uniformity and performance reliability of welded joints in thick-walled duplex steel components.

[0009] The above description is merely an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, specific embodiments of this application are given below. Attached Figure Description

[0010] To more clearly illustrate the technical solutions in this disclosure 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 merely exemplary. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0011] Figure 1 A schematic flowchart of a segmented thermally controlled welding method for thick-walled duplex steel components provided in an embodiment of this application; Figure 2 This is a schematic diagram of the segmented thermally controlled welding system for thick-walled duplex steel components provided in an embodiment of this application.

[0012] Figure labeling: 11 segmented unit planning and construction module, 12 multi-source sensing acquisition module, 13 thermal cycle feature extraction module, 14 welding process constraint generation module, 15 segmented thermal input deviation calculation module, 16 process parameter adjustment instruction generation module, 17 model optimization and update module. Detailed Implementation

[0013] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0014] Example 1: The segmented controlled-heat welding method for thick-walled duplex steel components provided in this disclosure is referred to below. Figure 1 The methods include: S1: Based on the geometric information of the welding area of ​​the duplex steel thick-walled component, divide it into welding segment units, and construct the segment unit planning set by combining the segment unit process constraints; Furthermore, step S1 also includes: Geometric information of the area to be welded is obtained through 3D laser scanning equipment and CAD models of components, including thickness distribution, bevel angle and weld centerline trajectory. Continuous sampling points are evenly distributed along the centerline trajectory of the weld, and the spatial coordinates, local plate thickness and bevel angle of each sampling point are collected to form a geometric information set of the component; Spatial discretization of the weld seam is performed based on the component geometric information set, dividing the weldable area of ​​the dual-phase steel thick-walled component into a series of continuous welding segment units; Obtain the component geometric information set, calculate the reference heat input of each welding segment unit based on the material performance manual and previous process experience, and set the upper and lower limits of heat input for each welding segment unit in conjunction with the process evaluation report to generate the segment unit process constraint set. Optimize the welding segment units based on the correlation between the baseline heat input and the upper and lower limits of heat input, and update the process constraint set of the segment units; The adjusted welding segment unit and the segment unit process constraint set are integrated to generate the segment unit planning set.

[0015] Specifically, the purpose of this step is to establish a spatial foundation for subsequent fine-grained control. First, complete geometric information of the area to be welded is obtained using a 3D laser scanning device and the component's CAD model, including thickness distribution, bevel angle, and weld centerline trajectory. All geometric information, after coordinate registration and point cloud denoising, is uniformly mapped to the weld reference coordinate system, providing a standardized spatial foundation for segmented analysis. Based on this, continuous sampling points are evenly distributed along the weld centerline trajectory, collecting the spatial coordinates, local plate thickness, and bevel angle of each sampling point. The sampling interval is set according to the weld length and process requirements, typically ranging from 5mm to 20mm to ensure a balance between accuracy and computational load. Finally, through the above spatial sampling, coordinate system unification processing, and geometric feature extraction, a component geometric information set oriented towards the weld is formed.

[0016] After obtaining the geometric information set of the component, the weld is spatially discretized according to the principle of geometric homogeneity. The discretization process is based on local plate thickness variations, bevel shape variations, and weld direction changes. Specifically, along the welding direction (longitudinal), segment boundaries are set at locations of abrupt thickness changes, bevel changes, or weld turns. Within each segment (transverse), the average thickness and representative bevel angle are used to characterize its cross-section. Finally, the welding area of ​​the duplex steel thick-walled component is divided into a series of continuous welding segment units along the weld direction. The geometric features within each segment unit are relatively uniform, and the geometric information contained in each segment unit includes: segment unit start coordinates, segment unit end coordinates, segment unit length, segment unit average thickness, and segment unit representative bevel angle.

[0017] To ensure that each segment unit is rationally planned and has a clear heat input control benchmark, a segment unit process constraint set is further constructed for each welding segment unit. This process constraint set includes: upper limit of heat input, lower limit of heat input, and benchmark heat input amount.

[0018] The upper limit of heat input is used to avoid excessive heat input leading to grain coarsening, performance degradation in the heat-affected zone, and accumulation of welding deformation. Its value is determined by three constraints: the continuous cooling transition (CCT) curve requirement, the constraint of structural thickness and component heat capacity on heat diffusion, and the requirements for controlling welding deformation and residual stress. The CCT curve is typically given as the maximum permissible line energy in the process qualification report (WPS / PQR). The constraint of structural thickness and component heat capacity on heat diffusion is obtained through a structural thickness function in the engineering. The requirements for controlling welding deformation and residual stress are obtained from the process qualification report (WPS / PQR) and structural manufacturing specifications. The minimum value of all constraints is ultimately taken as the upper limit of heat input for the segmented unit. The lower limit of heat input is used to avoid insufficient fusion, incomplete penetration, and increased cold cracking tendency due to insufficient heat input. This lower limit is determined by three constraints: the minimum effective line energy recorded in the process qualification report (WPS / PQR), the required preheating temperature, and forming quality constraints. The required preheating temperature is specified by material standards (such as ISO 2560) and process qualification reports (WPS / PQR), while forming quality constraints are obtained through process standards. The maximum value among these constraints is taken as the lower limit of the heat input for each segment unit. The upper and lower limits of the heat input obtained in this way constitute the executable energy input boundary for each segment unit, used to constrain the effective heat input range of each segment.

[0019] The baseline heat input is calculated based on a simplified heat balance model, and the specific calculation formula is as follows: Q ref =k×h avg ×L seg ×ΔT; Among them, Q ref The reference heat input is represented by k, which represents the thermal conductivity of the material. k is an inherent physical property parameter of the welding material and can be obtained directly from material performance manuals, material suppliers, or material technical specifications provided by steel mills. h avg L represents the average thickness of the segmented unit, derived from the geometric information of the welded segmented unit. seg ΔT represents the length of the segment unit, which is calculated based on the start and end positions of the segment and is also derived from the geometric information of the welding segment unit. ΔT represents the design temperature rise, which is the critical temperature difference required to heat the weld area to meet the requirements of fusion and microstructure. Its value is determined by a combination of material phase transformation temperature, preheating conditions and previous process experience.

[0020] The segmented units are optimized based on the correlation between the baseline heat input and the upper and lower limits of the heat input. The calculated baseline heat input must fall within the upper and lower limits of the heat input in the segmented process constraint set. If the calculation result exceeds the allowable range, the segmented unit is marked as a thermal risk segmented unit. For segmented units marked as thermal risk segmented units, segment refinement will be automatically triggered, re-dividing them into shorter segments. Generally, they will be uniformly divided into two segmented units to ensure that the thermal control of the segmented units is physically feasible. The segmented unit process constraint set is then updated based on the re-divided welding segmented units.

[0021] Finally, the adjusted welding segment units and segment unit process constraint sets are integrated to form a segment unit planning set. This dataset includes the segment unit number, segment unit start coordinates, segment unit end coordinates, segment unit length, segment unit average thickness, segment unit representative bevel angle, upper limit of heat input, lower limit of heat input, and reference heat input information.

[0022] S2: Deploy a sensor acquisition system to collect temperature information and welding process parameters of all welding segment units in real time, and generate a multi-source synchronous dataset of segment units. The sensor acquisition system includes an infrared temperature measurement array and a thermocouple sensor array. The temperature information includes surface temperature and interlayer temperature. Furthermore, step S2 also includes: Infrared temperature measurement arrays and thermocouple sensor arrays are deployed at key locations in each welding segment unit to obtain the surface temperature and interlayer temperature during welding. Welding process parameters, including welding current, welding voltage, and welding speed, are obtained from the welding power source and motion control system. The surface temperature, interpass temperature, welding current, welding voltage, and welding speed data are reorganized according to the welding segment units to construct a multi-source synchronous dataset for the segment units.

[0023] Specifically, after completing the segmentation into units, this step focuses on real-time acquisition of multi-source process data. By synchronously sensing surface temperature, interpass temperature, and welding process parameters, it provides complete and time-consistent foundational data for subsequent extraction of thermal cycle characteristic parameters and calculation of thermal input deviations. To achieve high temporal resolution and full-coverage process sensing, this step involves three aspects: sensor deployment, data acquisition methods, and synchronization mechanisms.

[0024] First, according to the segmented units planned in step S1, infrared temperature measurement arrays and thermocouple sensor arrays are deployed at key positions in each segmented unit to obtain the surface temperature and interpass temperature during welding. At the same time, welding process parameters are obtained based on the welding power source and motion control system.

[0025] Infrared thermography arrays perform non-contact measurement of weld surface temperature. Their placement follows the weld's direction, arranged laterally along the weld centerline to ensure the entire welding area remains within the array's scanning field of view. Specifically, the infrared thermography array is fixed to a rigid bracket next to the welding head, typically positioned 200-400mm above the weld to avoid interfering with the welding flame or molten pool while maintaining sufficient field of view coverage. The infrared thermography array acquires the surface temperature distribution of the weld and its heat-affected zone using a high frame rate scanning method, typically between 50-200 Hz. The sensor internally converts infrared radiation into temperature information to obtain the surface temperature. For each surface temperature, a timestamp, array number, and corresponding welding position are recorded simultaneously, ensuring clear spatial and temporal reference integrity for the surface temperature data.

[0026] Thermocouple sensor arrays are used to acquire temperature information of the internal regions during the welding process. Unlike infrared thermometry, thermocouples directly contact the workpiece being welded, allowing for precise measurement of temperature changes between weld layers. During welding, thermocouples are mechanically fixed or spot-welded to designated temperature measurement points near the weld seam, and their placement is closely related to the welding machine method. For example, if the welding method is single-sided welding, thermocouples are placed at the root of the back of the weld seam; if the welding method is multi-pass welding, thermocouples are fixed between passes n and n+1; if the welding method is thick plate welding, multiple thermocouples are placed along the thickness direction. The thermocouples output thermoelectric potential signals, which are converted into real-time interlayer temperatures after being processed by a cold junction compensation module and a high-precision acquisition card. The acquisition frequency is typically 10-50 Hz to capture the slow trend of temperature changes in the weld layers. Each thermocouple has a unique number, and its acquired data is marked as interlayer temperature data, along with a corresponding timestamp and welding position, facilitating subsequent joint analysis with surface temperature.

[0027] Aside from temperature information, welding current, welding voltage, and welding speed are fundamental process parameters in the welding process, essential data for calculating actual heat input, and also require real-time acquisition. Welding current and voltage are obtained directly from the welding power supply's internal real-time sampling values ​​via digital signal interfaces such as CAN bus, EtherCAT, and MODBUS-TCP. Because the sampling accuracy inside the welding machine is higher than that of external sensors, this digital reading method offers higher reliability. Welding speed is obtained through encoder feedback from the motion controller. The encoder measures the displacement change of the welding head and calculates the instantaneous speed based on the sampling interval. The motion controller typically provides real-time position data; by reading the displacement sequence output by the controller, the actual welding speed at each moment can be calculated.

[0028] During the real-time acquisition of surface temperature, interpass temperature, and welding process parameters, time synchronization of all acquisition modules is required to ensure that all data are aligned on the same timeline. Specific implementation steps include: distributing a high-precision synchronization clock to all acquisition modules via a timestamp server to ensure a time error of less than 1ms; immediately timestamping data upon arrival at the acquisition port; merging multiple source data into the same time series at the same time; and interpolating or aligning data with different frame rates to ensure time consistency.

[0029] After synchronizing the surface temperature data, interpass temperature data, and process parameter data, all data are reorganized into segmented units to construct a segmented unit multi-source synchronized dataset. This dataset serves as the sole original basis for subsequent thermal cycle calculations, deviation assessments, and process compensation. The dataset includes: surface temperature sequence, interpass temperature sequence, welding current sequence, welding voltage sequence, welding speed sequence, and the timestamp, welding position, and corresponding segmented unit number for each data point.

[0030] S3: Preprocess the temperature information in the segmented unit multi-source synchronous dataset, extract thermal cycle feature parameters, and construct a welding thermal cycle feature parameter set; Furthermore, step S3 also includes: Obtain multi-source synchronous datasets for segmented units, and perform noise reduction processing on surface temperature and interlayer temperature data; Surface transient temperature sequences and interlayer transient temperature sequences are extracted based on surface temperature and interlayer temperature data; Surface cooling rate and temperature gradient were calculated using the transient surface temperature sequence. Interlayer cooling rate and interlayer temperature deviation were calculated using the interlayer transient temperature sequence. The surface transient temperature sequence, interlayer transient temperature sequence, surface cooling rate, interlayer cooling rate, temperature gradient, and interlayer temperature deviation are reorganized according to a unified timestamp and welding segment unit number, while spatial information is retained according to the measurement point location, to construct a set of welding thermal cycle characteristic parameters.

[0031] Specifically, after constructing the segmented unit multi-source synchronous dataset, this step aims to obtain thermal cycling characteristic parameters with strong descriptiveness and clear physical meaning. Through systematic signal processing of temperature data, the raw data is gradually transformed into key characteristic parameters that can be used to judge thermal input deviation.

[0032] First, noise reduction is performed on the two types of temperature data in the segmented unit multi-source synchronous dataset. Due to its high frame rate and significant radiative disturbance, the surface temperature sequence is smoothed using an adaptive filtering algorithm to maintain responsiveness during rapid changes and effectively suppress high-frequency noise during stable periods. Thermocouple sampling frequencies are relatively low, so low-pass filtering is used to remove power frequency and random interference from the interlayer temperature sequence.

[0033] After obtaining two types of reliable temperature data, characteristic parameters reflecting the welding thermal cycle behavior are extracted from them. First, the surface transient temperature sequence and interpass transient temperature sequence of the segmented unit are obtained directly from the filtered surface temperature sequence and interpass temperature sequence. These two sequences record the temperature state of the surface and interior of the weld zone over time, respectively, and are the basic information for describing the thermal cycle.

[0034] Subsequently, the cooling rate is calculated by differentiating the slope of temperature change over time, yielding the surface cooling rate and interlayer cooling rate. These rates are obtained by differentiating the surface temperature-time curve and the interlayer temperature-time curve, respectively, and are used to characterize the rate of heat release during post-weld cooling. The two rates are automatically calculated point-by-point by the system, forming a continuous sequence on the time axis.

[0035] To determine the spatial temperature distribution variations in the weld area, it is necessary to calculate the temperature gradient based on multi-point temperature measurements taken by an infrared array at the same time. This reveals the direction and non-uniformity of heat diffusion across the weld surface. The specific calculation formula is as follows: ; in, This represents the temperature gradient between positions i and j at time t, used to describe the intensity of heat conduction. This represents the transient temperature of the infrared array at position i on the surface. This represents the transient temperature of the infrared array at position j on the surface. The distance between positions i and j can be obtained from the coordinates of the two positions.

[0036] Interlayer temperature deviation, obtained from interlayer temperatures collected by thermocouple sensors, describes the differences in heat accumulation within the material and measures the uniformity of heat transfer along the thickness direction. Its physical significance lies in revealing the inconsistency of thermal cycling along the depth direction. Based on the temperatures obtained by thermocouples at different interlayer locations, the interlayer temperature deviation is calculated by determining the temperature difference between different depths at the same time point. Its form can be expressed as: ; in, This represents the interlayer temperature deviation at time t. These represent the highest and lowest interlayer temperatures at time t, respectively, at the internal measuring points. The data comes from thermocouples placed at different interlayer locations. It should be noted that if the welding method is single-sided welding, the highest interlayer temperature is directly replaced by the surface temperature, while the lowest interlayer temperature still comes from the temperature data collected by the thermocouples. For other welding methods, the highest and lowest interlayer temperatures are calculated based on the transient interlayer temperature sequence.

[0037] After calculating the aforementioned thermal cycle characteristics, the surface transient temperature sequence, interlayer transient temperature sequence, surface cooling rate, interlayer cooling rate, temperature gradient, and interlayer temperature deviation need to be reorganized according to a unified timestamp and segment numbering. Spatial information is also retained based on the measurement point location to form a structured feature set. Each segment unit corresponds to a set of thermal cycle characteristic parameters that change continuously over time. All characteristic parameters are strictly aligned to the same time axis to ensure consistent temporal correlation between different physical quantities. Finally, all characteristic parameters are integrated in segment units to construct a welding thermal cycle characteristic parameter set. This parameter set comprehensively reflects the thermodynamic changes in the welding process across time, space, and depth dimensions, serving as the core basis for subsequent thermal input deviation identification.

[0038] S4: Based on material properties and welding procedure qualification reports, obtain the planned process parameters, preset line energy and thermal cycle characteristic parameter ranges during the welding process, and generate a welding control constraint set; Furthermore, step S4 also includes: Based on the material property database and welding procedure qualification report, the planned process parameters for each welding segment unit are obtained, including the preset welding current, preset welding voltage and preset welding speed. The preset line energy is calculated based on the planned process parameters, and the specific formula is as follows: ; Among them, Q set I represents the preset line energy of the welding segment unit. set U set V set These represent the preset welding current, preset welding voltage, and preset welding speed, respectively. Based on the thermal properties of the material and the welding metallurgical behavior, the range of thermal cycle characteristic parameters is set, specifically including the upper limit of surface temperature, the lower limit of surface temperature, the upper limit of interpass temperature, the lower limit of interpass temperature, as well as the allowable range of surface cooling rate, interpass cooling rate, temperature gradient, and interpass temperature deviation. The planned process parameters, preset line energy, and thermal cycle characteristic parameter ranges are uniformly integrated according to the welding segment unit number to construct a welding control constraint set.

[0039] Specifically, after obtaining the thermal cycle characteristic parameters of the segmented units, it is necessary to further construct a control benchmark and process constraint system for real-time judgment of welding heat input deviations. The goal of this step is to limit the range of actually executable process parameters, forming a control reference value system that runs through the entire process, so that subsequent deviation diagnosis can be carried out within a clearly permissible range.

[0040] First, based on the material property database, welding procedure qualification report, and specific welding specifications, the planned process parameters for each segment unit are obtained. According to the welding current, welding voltage, and welding speed recommended in the procedure qualification report, the segment-preset welding current I is set within the corresponding segment unit. set Preset welding voltage U set With preset welding speed V set The above parameters are determined by manual process planning and are derived from verified process qualification reports. They represent a stable and reliable basic process combination for welding projects.

[0041] After setting the planned process parameters, a preset line energy needs to be constructed for comparing heat input deviations. Based on the physical definition of welding line energy, the preset line energy for each segment can be calculated, in the form of: ; Among them, Q set The preset linear energy representing the segmented unit is a core indicator used to judge heat input deviation. set U set V set These are the preset welding current, preset welding voltage, and preset welding speed, respectively. This formula reflects the combined effect of energy input and travel speed.

[0042] At the same time, it is also necessary to establish process safety boundaries that affect welding quality based on the thermal properties of the materials and the welding metallurgical behavior. Specifically, this includes the upper limit of surface temperature, the lower limit of surface temperature, the upper limit of interpass temperature, the lower limit of interpass temperature, as well as the allowable ranges for surface cooling rate, interpass cooling rate, temperature gradient, and interpass temperature deviation.

[0043] The upper limit of surface temperature is set based on the continuous cooling transition curve (CCT) of the material, the highest surface temperature corresponding to the qualified joint in the process evaluation report, and the requirements for oxidation resistance and deformation prevention of the component surface. For example, the upper limit of surface temperature for Q345 and Q355 steel is approximately 550–650°C.

[0044] The lower limit of surface temperature is usually obtained from the recommended preheating temperature and the minimum temperature requirement to prevent cold cracking, such as 80-120°C for ordinary steel.

[0045] Interpass temperature must be within a strictly permissible range to avoid the risk of cold cracking due to excessively low interpass temperatures, or metallurgical problems such as grain coarsening and decreased toughness due to excessively high interpass temperatures. Therefore, the upper and lower limits of the interpass temperature are determined based on the material's continuous cooling transformation (CCT) curve, welding procedure qualification test results, and applicable structural steel standards. For example, for high-strength low-alloy steels, the interpass temperature is typically controlled within the range of 100-250°C.

[0046] Similarly, the allowable ranges for surface cooling rate, interlayer cooling rate, temperature gradient, and interlayer temperature deviation are determined based on material toughness requirements and established process evaluation data.

[0047] The upper limits for surface cooling rate and interpass cooling rate are derived from the requirement to ensure that the weld metal and heat-affected zone do not enter the high-hardness martensite region. Their values ​​are determined based on the transformation boundary from the pearlite / bainite region to the martensite region in the CCT curve. The lower limits for surface cooling rate and interpass cooling rate are set based on the risk of excessive heat input leading to grain coarsening or a decrease in the toughness of the heat-affected zone.

[0048] The allowable ranges for temperature gradients and interpass temperature deviations are derived from the requirements for controlling residual stress in structural steel welding and the uniformity of microstructure in the heat-affected zone. These ranges are generally determined through a combination of trial welding measurements, thermal simulation experiments, and finite element thermo-stress analysis. For example, in some shipbuilding thick plate welding scenarios, the allowable interpass temperature deviation can be controlled within the range of 20-40°C. This range is derived from the statistical results of thermal simulation experiments on the uniformity of microstructure distribution. Temperature gradients are generally required to be less than 15-30°C / mm to prevent high gradients in localized areas of thick-walled structures that could lead to hot cracking.

[0049] All the above limits are determined comprehensively based on the mechanical and metallurgical properties of the material, the type of weld joint, the plate thickness, and the results of the welding procedure qualification; there are no arbitrary settings. For indicators that are not given explicit values ​​in the standard, feasible ranges are obtained through verified welding procedure qualification reports, historical weld quality databases of similar products, and thermal simulation tests, and are calibrated according to the target component during system deployment.

[0050] After calculating and collecting the preset line energy and various process boundary parameters, all parameters need to be integrated according to the segment unit number, so that each segment unit forms a complete control reference structure. Each segment unit includes its corresponding preset welding current, preset welding voltage, preset welding speed, preset line energy, upper limit of surface temperature, lower limit of surface temperature, upper limit of interpass temperature, lower limit of interpass temperature, as well as the allowable range of surface cooling rate, interpass cooling rate, temperature gradient, and interpass temperature deviation. Through this structured organization, the system can accurately reference the process reference of the corresponding segment in subsequent real-time calculations.

[0051] Ultimately, the welding control constraint set constructed through this step includes target benchmark values ​​obtained based on actual process evaluation and constraint conditions determined based on material properties and process specifications. This control benchmark system is fully aligned with the thermal cycling characteristic parameters obtained in step S3 above in terms of time and segmented structure, providing a rigorous basis for judging thermal input deviation identification and adaptive control logic.

[0052] S5: Based on the segmented unit multi-source synchronous dataset, welding thermal cycle characteristic parameter set and welding control constraint set, calculate the comprehensive deviation of the thermal input of all welding segmented units and obtain the segmented thermal input deviation instruction set; Furthermore, step S5 also includes: Obtain the set of characteristic parameters of the welding thermal cycle and the set of welding control constraints, and compare the values ​​of the characteristic parameters of the thermal cycle with the range of the characteristic parameters of the thermal cycle. If the value of a thermal cycle characteristic parameter is found to exceed the upper limit of the interval, calculate the deviation corresponding to that thermal cycle characteristic parameter. The specific calculation formula is as follows: ; in, This represents the deviation of the thermal cycle characteristic parameter i. The measured values ​​of this thermal cycle characteristic parameter are obtained from the welding thermal cycle characteristic parameter set. and These are the upper and lower limits of the allowable range corresponding to the characteristic parameters of the thermal cycle, obtained through the welding control constraint set; If the value of a thermal cycle characteristic parameter is found to be lower than the lower limit of the interval, the formula for calculating the deviation of that thermal cycle characteristic parameter can be expressed as: ; If the value of the thermal cycle characteristic parameter is found to be within the range, the deviation corresponding to the thermal cycle characteristic parameter is assigned to zero. The actual linear energy Q is calculated based on the welding current, welding voltage, and welding speed collected centrally from the multi-source synchronous data of the segmented unit. act The specific calculation formula is as follows: ; Among them, I act U act V act These represent the welding current, welding voltage, and welding speed centrally acquired from the multi-source synchronous dataset of the segmented unit, respectively. The actual line energy is compared with the preset line energy to obtain the path tracking deviation. The specific calculation formula is as follows: ; in, Q represents the path tracking deviation.act Q represents the actual linear energy. set Represents the preset line energy; The deviations corresponding to all thermal cycle characteristic parameters are weighted and fused with the path tracking deviation to obtain a comprehensive deviation characterizing the welding heat input state. The specific calculation formula is as follows: ; in, This represents the comprehensive deviation of the heat input of the segmented unit at the current moment; it is a dimensionless heat input deviation index. The weights corresponding to the characteristic parameters of the i-th thermal cycle are: This represents the deviation corresponding to the characteristic parameter of the i-th thermal cycle. The weights representing path tracking deviations This represents the path tracking deviation. The comprehensive deviation of each welding segment unit, which is continuously updated throughout the welding process, is organized in chronological order to form a real-time segmented thermal input deviation instruction set.

[0053] Specifically, after obtaining the multi-source synchronous dataset of segmented units, the set of welding thermal cycle characteristic parameters, and the set of welding control constraints, it is necessary to construct a deviation decision-making mechanism capable of evaluating the welding thermal input state in real time. This step involves comparing the real-time information obtained in S2 and S3 with the control benchmark set in S4 item by item, and combining this with the actual linear energy calculated in real time to form a comprehensive deviation quantity that can be directly used for process adjustment, thus achieving a quantitative assessment of the thermal input state. The goal of this step is to unify data from different sources and with different physical meanings into a calculable deviation framework, ensuring that the thermal input state of each segmented unit can be accurately characterized.

[0054] First, the thermal cycling characteristic parameters obtained in step S3 are compared with the corresponding allowable ranges set in step S4 for compliance. For each segmented unit, at each time step, it is sequentially checked whether the transient surface temperature is within the allowable surface temperature range and whether the transient interlayer temperature is within the allowable interlayer temperature range. Similarly, it is determined whether the surface cooling rate, interlayer cooling rate, temperature gradient, and interlayer temperature deviation fall within the allowable ranges for surface cooling rate, interlayer cooling rate, temperature gradient, and interlayer temperature deviation specified for that segment. These allowable ranges are all derived from the process safety boundaries constructed in step S4 based on material properties and process specifications, and are rigid constraints for welding process quality control.

[0055] For any of the above real-time characteristic parameters, if its instantaneous measured value exceeds the upper or lower limit of the interval, the deviation of that characteristic parameter needs to be calculated to characterize the degree of deviation. The deviation is calculated using a normalized method, allowing physical quantities with different dimensions to be compared within the same evaluation system. For example, when the measured value is higher than the upper limit of the interval, the deviation can be expressed as: ; in, The deviation representing characteristic parameter i mainly involves surface transient temperature, interlayer transient temperature, surface cooling rate, interlayer cooling rate, temperature gradient, and interlayer temperature deviation. The current measured value of this characteristic parameter is obtained from the welding thermal cycle characteristic parameter set. and These are the upper and lower limits of the allowable interval corresponding to the feature parameter, obtained through the welding control constraint set.

[0056] Similarly, when the measured value is lower than the lower limit of the interval, the deviation can be expressed as: ; The final calculated deviation reflects the proportion of the feature parameter exceeding the allowable range within the dimensionless interval; the larger the absolute value, the more severe the deviation. This step calculates the deviation for all feature parameters exceeding the allowable range in this way, and assigns a deviation of zero to feature parameters that do not exceed the range, so that they do not affect subsequent comprehensive decision-making.

[0057] In addition to determining the range deviation of various thermal cycle characteristics, it is also necessary to calculate the path tracking deviation of the welding process, which describes the degree of deviation of the actual line energy from the preset line energy. The actual line energy Q is calculated based on the welding current, welding voltage, and welding speed collected in real time from the multi-source synchronous data of the segmented unit in step S2. act And the preset line energy Q in the welding control constraint set of step S4. set Perform a comparison. Path tracking deviation can be expressed as: ; in, Q is the path tracking deviation, used to measure whether the current energy input deviates from the preset energy trajectory. It is an important indicator reflecting the accuracy of heat input control. act The formula for obtaining Q is the same. set Similarly, the calculation form can be expressed as Among them, I act U act V act These are the welding current, welding voltage, and welding speed obtained from the multi-source synchronous dataset of the segmented unit, respectively.

[0058] After obtaining the deviations of various characteristic parameters and path tracking deviations, they need to be weighted and fused under a unified decision-making framework to obtain a comprehensive deviation that characterizes the welding heat input state. To this end, weight coefficients are assigned to different parameters, enabling the system to determine the priority of correction based on their metallurgical sensitivity and quality risk level. For example, features directly related to microstructure transformation, such as interpass temperature and cooling rate, typically have higher weights than temperature gradients or surface temperatures. The comprehensive deviation is calculated as follows: ; in, This represents the comprehensive deviation of the segmented unit at the current moment, and is a dimensionless thermal input deviation index. The weights corresponding to the characteristic parameters of the i-th thermal cycle are: The deviation of characteristic parameter i is represented. As the weight of the path tracking deviation, This represents the path tracking deviation. A positive value indicates that the current heat input is too high, and an adjustment strategy to reduce the heat input should be adopted; a negative value indicates that the heat input is too low.

[0059] Specifically, when constructing the calculation model for the comprehensive deviation, it is necessary to assign initial weight coefficients to various thermal cycle characteristic parameters so that the control model can respond differently based on the degree of influence of different characteristic deviations on welding quality. Specifically, thermal cycle characteristic parameters such as surface temperature, interpass temperature, surface cooling rate, interpass cooling rate, temperature gradient, and interpass temperature deviation constitute the key physical quantities that determine the welding thermal history. They represent the transient heat input state of the weld region, the accumulated heat inside the thick plate, the transformation rate of the material structure, and the uniformity of heat distribution during the welding process, respectively. Based on the differences in the importance of these physical quantities, parameters directly related to metallurgical quality and whose deviations may lead to cracks, hardened structures, or significant performance degradation are given higher weights when setting the weights. For example, interpass temperature and interpass cooling rate are core indicators affecting the microstructure transformation behavior and cold crack resistance of welded joints. Deviations in these parameters often carry significant metallurgical risks, and therefore are typically assigned the highest weight in the weighting system, with a weight range of 0.20–0.30. Temperature gradient reflects the thermal inhomogeneity of thick-walled components, directly impacting residual stress levels and weld deformation. It represents a medium-level risk affecting structural integrity, and its weight is set between 0.10 and 0.15. While surface temperature and surface cooling rate are easier to monitor, their impact focuses more on weld formation stability and surface quality, exhibiting lower metallurgical sensitivity compared to interpass indicators. Therefore, their weight is generally set between 0.05 and 0.10. Interpass temperature deviation is a key quantitative indicator characterizing the uniformity of heat transfer and the gradient of the thermal field distribution during thick plate welding. It is highly correlated with the material's CCT curve and the allowable range of heat input variation. Its weight is adjusted appropriately based on material type, component thickness, and actual working conditions, typically ranging from 0.15 to 0.20.

[0060] In addition to the weights related to the physical quantities of the thermal cycle mentioned above, the model also needs to consider path tracking deviation, i.e., the degree of deviation between the actual linear energy and the preset linear energy. Path tracking deviation reflects the stability of the equipment's execution state and is a fundamental factor in achieving automated closed-loop control. However, its direct impact on the final microstructure and properties of the weld is weaker than that of metallurgically sensitive parameters such as interpass temperature and cooling rate. Therefore, as an auxiliary weight term, its weight is usually set in the range of 0.10 to 0.20. In the process of determining the weights, not only were the metallurgical characteristics of the welding materials, the process qualification report, and the corresponding CCT curves of the materials considered, but also historical welding data, weld quality inspection results, and the actual response capability of the control system were combined to calibrate the weights, so that the final weight system can take into account both metallurgical safety and the adjustability of the control process. The final determined set of weights satisfies a total value of 1, which can ensure that the total deviation has a stable and interpretable response behavior under different combinations of deviations.

[0061] After completing the above calculations, the comprehensive deviation of each segment unit, which is continuously updated throughout the welding process, needs to be organized in chronological order to form a real-time segmented thermal input deviation instruction set. This instruction set is the core input upon which the subsequent thermal input adaptive control module performs gain adjustment, speed correction, or current adjustment.

[0062] S6: Utilize the segmented thermal input deviation instruction set to calculate the final current adjustment and final speed adjustment, and generate welding process parameter adjustment instructions; Furthermore, step S6 also includes: The comprehensive deviation is mapped to the target energy correction, and the specific calculation formula is as follows: ; in, This represents the target energy correction amount. Represents the overall deviation. Represents the preset line energy. Represents the correction factor; The welding current, welding voltage, and welding speed are collected in real time from the multi-source synchronous dataset of the segmented unit. Two types of sensitivity coefficients, a and b, are calculated. The specific calculation formulas for the two coefficients are as follows: ; Where a is the current sensitivity coefficient. For welding voltage, For welding speed; ; Where b is the velocity sensitivity coefficient. For actual linear energy, For welding speed; Based on the linear constraint relationship, the theoretical current adjustment and theoretical speed adjustment are solved using the weighted quadratic cost minimization method. The specific formulas are as follows: ; ; in, Represents the theoretical current adjustment amount. Represents the theoretical speed adjustment amount. Represents the target energy correction amount, 'a' is the current sensitivity coefficient, and 'b' is the velocity sensitivity coefficient. , These are the weighting coefficients; The final current adjustment and final speed adjustment are obtained based on the theoretical current adjustment and theoretical speed adjustment. The specific calculation formula is as follows: ; ; in, This represents the final current adjustment amount. Represents the final speed adjustment amount. Represents the theoretical current adjustment amount. Represents the theoretical speed adjustment amount. This represents the piecewise gain coefficient, and the specific calculation formula is as follows: ,in Represents the reference gain coefficient. This represents the heat capacity per unit area of ​​the segment. The heat capacity per unit area of ​​the reference segment. Represents the index factor; Safety checks are performed based on the final current and speed adjustments, and welding process parameter adjustment instructions are generated.

[0063] Specifically, after calculating the heat input deviation, the segmented heat input deviation command obtained in step S5 needs to be converted into process parameter adjustment amounts that can be directly executed by the welding equipment. The core task of this step is to achieve a closed-loop mapping from deviation determination to equipment action, and to ensure that all adjustment actions are performed within the previously established process safety boundaries to avoid welding quality risks caused by over-adjustment.

[0064] After obtaining the comprehensive deviation output in step S5, the comprehensive deviation is first mapped to the target energy correction. The specific calculation formula is as follows: ; in, This represents the target energy correction amount. A positive value indicates that the line energy needs to be increased proportionally, while a negative value indicates that the line energy needs to be decreased. Represents the overall deviation. Represents the preset line energy. This represents the correction coefficient, which is dimensionless and usually less than or equal to 1. It is used to prevent excessive single-step adjustments and ensure the stability of the control system. Its initial value is set based on process experiments and can be set to 1 here.

[0065] Read the welding current, welding voltage, and welding speed collected in real time in step S2, and calculate the two types of sensitivity coefficients a and b. The specific calculation formulas for the two are as follows: ; Where 'a' is the current sensitivity coefficient, representing the rate of change of linear energy caused by a small change in welding current under the current operating conditions. For welding voltage, This refers to the welding speed.

[0066] ; Where b is the speed sensitivity coefficient, representing the rate of change of linear energy caused by a small change in welding speed under the current operating conditions. For actual linear energy, This refers to the welding speed.

[0067] After calculating the two types of sensitivity coefficients, in order to simultaneously adjust the current and velocity to achieve target energy correction, it is necessary to apply the linear constraint relationship. The theoretical current adjustment and theoretical speed adjustment are solved using the weighted quadratic cost minimization method, and the specific formulas are as follows: ; ; in, Represents the theoretical current adjustment amount. Represents the theoretical speed adjustment amount. Represents the target energy correction amount, 'a' is the current sensitivity coefficient, and 'b' is the velocity sensitivity coefficient. , These are the weighting coefficients.

[0068] Specific explanation of weighting coefficients , Based on the equipment's response capability, the process's sensitivity to changes in current and speed, and safety strategy settings, calibration is primarily performed based on the equipment's permissible single-step adjustment range and initial process stability. The specific formula is as follows: , ,in , These are the nominal single-step changes in current and speed, representing the reference step size that will not significantly compromise welding stability, provided by the process qualification or equipment manual. The preset process parameters are derived from the preset welding current and preset welding speed in step S4. This standardized form enables... , It is dimensionless and matches the process dimensions.

[0069] Considering the differences in thermal inertia caused by the different geometric dimensions of the various segments, a dynamic gain proportional to the heat capacity of each segment is applied to the theoretical adjustments of the current and velocity, so that the control response matches the thermophysical characteristics of the segments. First, the heat capacity per unit area of ​​each segment is defined. ,in Represents density, Specific heat capacity is represented by values ​​from a materials property database. The average thickness is derived from the segmented unit planning set obtained in step S1. Next, the segmented gain coefficient K is calculated based on the heat capacity per unit area of ​​the segments. seg The specific calculation formula is as follows: ,in This represents the reference gain coefficient calibrated through process testing; its initial value is typically set to 1. The heat capacity per unit area representing the reference segment is typically chosen from a segment with a representative thickness or the most common thickness. Its main function is to normalize the heat capacity per unit area. The index factor, typically between 0 and 1, needs to be determined through process experiments or simulations based on physical models. Its main function is to adjust the intensity of the thermal inertia effect. Finally, the final adjustment amount is calculated based on the piecewise gain coefficient. The specific calculation formula is as follows: ; ; in, This represents the final current adjustment amount. Represents the final speed adjustment amount. Represents the piecewise gain coefficient. Represents the theoretical current adjustment amount. This represents the theoretical speed adjustment amount.

[0070] After calculating the current and speed adjustments after segmented gain compensation, the system will perform a series of feedforward safety checks before execution to absolutely ensure the safety of the welding process and strictly follow the process specifications.

[0071] First, based on the final adjustment amount calculated so far, the new parameter values ​​after execution are predicted, including the new welding current and welding speed. Then, the predicted line energy is calculated based on the new welding current, welding speed, and welding voltage, and a safety check is performed according to two constraint principles. The first is energy boundary check, which determines whether the predicted line energy exceeds the process allowable range determined in step S1 for this welding segment unit. If it exceeds the upper limit, it indicates that the adjustment will lead to excessive heat input, and the system will proportionally reduce the final current adjustment and final speed adjustment until the predicted line energy equals the maximum allowable value; if it is less, it will be adjusted proportionally until the predicted line energy equals the minimum allowable value. The second is single-parameter limit check, which, based on meeting the energy constraints, further checks whether the new welding current and welding speed are within the capacity range of the welding equipment itself and the safety limits specified in the process procedure. If any parameter exceeds the limit, it is forcibly truncated to the corresponding limit value, and another parameter is recalculated to ensure that the target energy correction requirement is approached as close as possible while meeting the limit values.

[0072] After all verifications are completed, the final welding process parameter adjustment command is generated. This command includes the final current and speed adjustments, which are processed with safety truncation and can be executed immediately. This command is synchronously sent to the welding power supply and motion controller via a real-time communication interface, driving the equipment to complete precise process parameter adjustments. Simultaneously, all key data from this adjustment is written to the segmented thermal input correction record for subsequent weld quality traceability, process optimization, and adaptive model updates.

[0073] S7: Based on the actual weld microstructure and properties during long-term welding, a parameter optimization algorithm is used to optimize and adjust the process, thereby achieving continuous updating of the segmented heat-controlled welding model.

[0074] Specifically, after completing real-time control and deviation correction of the welding process, this step aims to systematically and automatically calibrate the coefficients involved in the control model based on a large amount of historical data accumulated in actual production operations. This enables the control system to achieve performance optimization, adaptive evolution, and continuous adaptability to process changes during long-term operation. This step establishes a causal relationship model between regulatory behavior and welding results, enabling the system to possess higher structural stability and thermal input control accuracy in subsequent welding tasks.

[0075] First, the segmented heat input correction records are collected and organized from step S6. These records detail the final adjustment amounts for each segment unit, providing complete time-series data describing the system's adjustment behavior to the welding physical state. Simultaneously, after the welding task is completed, post-weld microstructure inspection and quality assessment methods (including metallographic observation, hardness distribution measurement, and, if necessary, impact toughness testing) are used to obtain a dataset of the microstructure properties of the weld and heat-affected zone, which characterizes the actual welding quality output.

[0076] After obtaining the two types of data mentioned above, a correlation model between adjustment behavior and welding results is established to achieve quantitative optimization of the control strategy. Specifically, a supervised machine learning algorithm is used to construct a mapping relationship model. The heat input adjustment behavior, real-time temperature evolution characteristics, deviation magnitude, and final microstructure performance indicators of each segment unit during welding are used as training samples. Through regression modeling and feature contribution analysis of a large number of historical samples, the influence of control parameters on welding quality is identified, and the sensitivity of different deviation types to the final microstructure stability is clarified. This modeling process can reveal which deviation terms are oversensitive or lack response, thus providing a basis for optimizing the weighting coefficients.

[0077] Based on the modeling results above, the system further optimizes various coefficients in steps S5 and S6, such as the weighting coefficient in step S5 and the correction coefficient and piecewise gain coefficient in step S6. The optimization principle aims to maximize the stability of weld microstructure properties, enabling the control system to improve the response strength to key deviation types (such as interpass temperature deviation and cooling rate deviation) and reduce over-adjustment of minor deviation terms, while ensuring that the process safety boundaries are not exceeded. Furthermore, the system automatically corrects and adjusts the gain based on the statistical characteristics of the actual welding scenario, making the control output more robust, reasonable, and interpretable. The optimized parameters are not immediately used for real-time control but undergo self-consistency verification by the validation module to ensure that the new parameters do not cause instability or overshoot in certain operating conditions.

[0078] After a systematic update of the parameters, the new coefficients are written into the parameter library of the control model to replace the old parameters. Thus, in the next welding task, the control system will use the updated parameters to perform deviation judgment and closed-loop adjustment, allowing it to continuously accumulate experience and gradually improve decision-making accuracy over long-term operation. This self-learning update mechanism based on historical data enables the control system to continuously evolve with changes in operating conditions, and is an indispensable key step in achieving intelligent, adaptive welding control.

[0079] Example 2: Based on the same inventive concept as the segmented heat-controlled welding method for thick-walled duplex steel components described in the foregoing examples, this application also provides a segmented heat-controlled welding system for thick-walled duplex steel components. Please refer to the appendix. Figure 2 The system includes: The segmented unit planning and construction module 11 is used to divide the welding segmented units according to the geometric information of the welding area of ​​the duplex steel thick-walled component, and to construct the segmented unit planning set in combination with the segmented unit process constraints. The multi-source sensor acquisition module 12 is used to deploy a sensor acquisition system to collect temperature information and welding process parameters of all welding segment units in real time and generate a multi-source synchronous dataset of segment units. The sensor acquisition system includes an infrared temperature measurement array and a thermocouple sensor array. The temperature information includes surface temperature and interlayer temperature. The thermal cycling feature extraction module 13 is used to preprocess the temperature information in the segmented unit multi-source synchronous dataset, extract thermal cycling feature parameters, and construct a welding thermal cycling feature parameter set. The welding process constraint generation module 14 is used to obtain the planned process parameters, preset line energy and thermal cycle characteristic parameter ranges during the welding process based on material properties and welding process qualification reports, and generate a welding control constraint set. The segmented thermal input deviation calculation module 15 is used to calculate the comprehensive deviation of the thermal input of all welding segment units based on the segmented unit multi-source synchronous dataset, the welding thermal cycle characteristic parameter set and the welding control constraint set, and to obtain the segmented thermal input deviation instruction set. The process parameter adjustment instruction generation module 16 is used to calculate the welding current adjustment amount and the welding speed adjustment amount using the segmented heat input deviation instruction set, and generate welding process parameter adjustment instructions. The model optimization and update module 17 is used to optimize and adjust the process by using parameter optimization algorithms based on the actual weld microstructure and properties during long-term welding, so as to realize the continuous updating of the segmented heat-controlled welding model.

[0080] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0081] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A segmented controlled-heat welding method for thick-walled duplex steel components, characterized in that, The method includes: Based on the geometric information of the welding area of ​​the duplex steel thick-walled component, welding segment units are divided, and a segment unit planning set is constructed by combining the process constraints of the segment units. A sensor acquisition system is deployed to collect temperature information and welding process parameters of all welding segment units in real time, and generate a multi-source synchronous dataset of segment units. The sensor acquisition system includes an infrared temperature measurement array and a thermocouple sensor array. The temperature information includes surface temperature and interlayer temperature. The temperature information in the segmented unit multi-source synchronous dataset is preprocessed to extract thermal cycle feature parameters, and a welding thermal cycle feature parameter set is constructed, including: Obtain multi-source synchronous datasets for segmented units, and perform noise reduction processing on surface temperature and interlayer temperature data; Surface transient temperature sequences and interlayer transient temperature sequences are extracted based on surface temperature and interlayer temperature data; Surface cooling rate and temperature gradient were calculated using the transient surface temperature sequence. Interlayer cooling rate and interlayer temperature deviation were calculated using the interlayer transient temperature sequence. The surface transient temperature sequence, interlayer transient temperature sequence, surface cooling rate, interlayer cooling rate, temperature gradient and interlayer temperature deviation are reorganized according to a unified timestamp and welding segment unit number, while spatial information is retained according to the measurement point location to construct a set of welding thermal cycle characteristic parameters. Based on material properties and welding procedure qualification reports, the planned process parameters, preset line energy and thermal cycle characteristic parameter ranges during the welding process are obtained, and a welding control constraint set is generated. Based on the segmented unit multi-source synchronous dataset, the welding thermal cycle characteristic parameter set and the welding control constraint set, the comprehensive deviation of the thermal input of all welding segmented units is calculated to obtain the segmented thermal input deviation instruction set. Using the segmented thermal input deviation instruction set, the final current adjustment and final speed adjustment are calculated, and welding process parameter adjustment instructions are generated. Based on the actual weld microstructure and properties during long-term welding, a parameter optimization algorithm is used to optimize and adjust the process, thereby achieving continuous updating of the segmented heat-controlled welding model.

2. The segmented controlled-heat welding method for thick-walled duplex steel components as described in claim 1, characterized in that, Construct a segmented unit planning set, including: Geometric information of the area to be welded is obtained through 3D laser scanning equipment and CAD models of components, including thickness distribution, bevel angle and weld centerline trajectory. Continuous sampling points are evenly distributed along the centerline trajectory of the weld, and the spatial coordinates, local plate thickness and bevel angle of each sampling point are collected to form a geometric information set of the component; Spatial discretization of the weld seam is performed based on the component geometric information set, dividing the weldable area of ​​the duplex steel thick-walled component into a series of continuous welding segment units; Obtain the component geometric information set, calculate the reference heat input of each welding segment unit based on the material performance manual and previous process experience, and set the upper and lower limits of heat input for each welding segment unit in conjunction with the process evaluation report to generate the segment unit process constraint set; Optimize the welding segment units based on the correlation between the baseline heat input and the upper and lower limits of heat input, and update the process constraint set of the segment units; The adjusted welding segment unit and the segment unit process constraint set are integrated to generate the segment unit planning set.

3. The segmented controlled-heat welding method for thick-walled duplex steel components as described in claim 1, characterized in that, Generate segmented unit multi-source synchronized datasets, including: Infrared temperature measurement arrays and thermocouple sensor arrays are deployed at key locations in each welding segment unit to obtain the surface temperature and interlayer temperature during welding. Welding process parameters, including welding current, welding voltage, and welding speed, are obtained from the welding power source and motion control system. The surface temperature, interpass temperature, welding current, welding voltage, and welding speed data are reorganized according to the welding segment units to construct a multi-source synchronous dataset for the segment units.

4. The segmented controlled-heat welding method for thick-walled duplex steel components as described in claim 1, characterized in that, Generate a set of welding control constraints, including: Based on the material property database and welding procedure qualification report, the planned process parameters for each welding segment unit are obtained, including the preset welding current, preset welding voltage and preset welding speed. The preset line energy is calculated based on the planned process parameters, and the specific formula is as follows: ； Among them, Q set I represents the preset line energy of the welding segment unit. set U set V set These represent the preset welding current, preset welding voltage, and preset welding speed, respectively. Based on the thermal properties of the material and the welding metallurgical behavior, the range of thermal cycle characteristic parameters is set, specifically including the upper limit of surface temperature, the lower limit of surface temperature, the upper limit of interpass temperature, the lower limit of interpass temperature, as well as the allowable range of surface cooling rate, interpass cooling rate, temperature gradient, and interpass temperature deviation. The planned process parameters, preset line energy, and thermal cycle characteristic parameter ranges are uniformly integrated according to the welding segment unit number to construct a welding control constraint set.

5. The segmented controlled-heat welding method for thick-walled duplex steel components as described in claim 1, characterized in that, Obtain the segmented thermal input deviation instruction set, including: Obtain the set of characteristic parameters of the welding thermal cycle and the set of welding control constraints, and compare the values ​​of the characteristic parameters of the thermal cycle with the range of the characteristic parameters of the thermal cycle. If the value of a thermal cycle characteristic parameter is found to exceed the upper limit of the interval, calculate the deviation corresponding to that thermal cycle characteristic parameter. The specific calculation formula is as follows: ； in, This represents the deviation of the thermal cycle characteristic parameter i. The measured values ​​of this thermal cycle characteristic parameter are obtained from the welding thermal cycle characteristic parameter set. and These are the upper and lower limits of the allowable range corresponding to the characteristic parameters of the thermal cycle, obtained through the welding control constraint set; If the value of a thermal cycle characteristic parameter is found to be lower than the lower limit of the interval, the formula for calculating the deviation of that thermal cycle characteristic parameter can be expressed as: ； If the value of the thermal cycle characteristic parameter is found to be within the range, the deviation corresponding to the thermal cycle characteristic parameter is assigned to zero. The actual linear energy Q is calculated based on the welding current, welding voltage, and welding speed collected centrally from the multi-source synchronous data of the segmented unit. act The specific calculation formula is as follows: ； Among them, I act U act V act These represent the welding current, welding voltage, and welding speed centrally acquired from the multi-source synchronous dataset of the segmented unit, respectively. The actual line energy is compared with the preset line energy to obtain the path tracking deviation. The specific calculation formula is as follows: ； in, Q represents the path tracking deviation. act Q represents the actual linear energy. set Represents the preset line energy; The deviations corresponding to all thermal cycle characteristic parameters are weighted and fused with the path tracking deviation to obtain a comprehensive deviation characterizing the welding heat input state. The specific calculation formula is as follows: ； in, This represents the comprehensive deviation of the heat input of the segmented unit at the current moment; it is a dimensionless heat input deviation index. The weights corresponding to the characteristic parameters of the i-th thermal cycle are: This represents the deviation corresponding to the characteristic parameter of the i-th thermal cycle. The weights representing path tracking deviations This represents the path tracking deviation. The comprehensive deviation of each welding segment unit, which is continuously updated throughout the welding process, is organized in chronological order to form a real-time segmented thermal input deviation instruction set.

6. The segmented controlled-heat welding method for thick-walled duplex steel components as described in claim 1, characterized in that, Generate welding process parameter adjustment instructions, including: The comprehensive deviation is mapped to the target energy correction, and the specific calculation formula is as follows: ； in, This represents the target energy correction amount. Represents the overall deviation. Represents the preset line energy. Represents the correction factor; The welding current, welding voltage, and welding speed are collected in real time from the multi-source synchronous dataset of the segmented unit. Two types of sensitivity coefficients, a and b, are calculated. The specific calculation formulas for the two coefficients are as follows: ； Where a is the current sensitivity coefficient. For welding voltage, For welding speed; ； Where b is the velocity sensitivity coefficient. For actual linear energy, For welding speed; Based on the linear constraint relationship, the theoretical current adjustment and theoretical speed adjustment are solved using the weighted quadratic cost minimization method. The specific formulas are as follows: ； ； in, Represents the theoretical current adjustment amount. Represents the theoretical speed adjustment amount. Represents the target energy correction amount, 'a' is the current sensitivity coefficient, and 'b' is the velocity sensitivity coefficient. , These are the weighting coefficients; The final current adjustment and final speed adjustment are obtained based on the theoretical current adjustment and theoretical speed adjustment. The specific calculation formula is as follows: ； ； in, Represents the final current adjustment amount. Represents the final speed adjustment amount. Represents the theoretical current adjustment amount. Represents the theoretical speed adjustment amount. This represents the piecewise gain coefficient, and the specific calculation formula is as follows: ,in Represents the reference gain coefficient. This represents the heat capacity per unit area of ​​the segment. The heat capacity per unit area of ​​the reference segment. Represents the index factor; Safety checks are performed based on the final current and speed adjustments, and welding process parameter adjustment instructions are generated.

7. A segmented heat-controlled welding system for thick-walled duplex steel components, characterized in that, The system is used to implement the segmented thermally controlled welding method for duplex steel thick-walled components as described in any one of claims 1 to 6, and the system comprises: The segmented unit planning and construction module is used to divide the welding segmented units according to the geometric information of the welding area of ​​the duplex steel thick-walled component, and construct the segmented unit planning set in combination with the segmented unit process constraints. A multi-source sensor acquisition module is used to deploy a sensor acquisition system to collect temperature information and welding process parameters of all welding segment units in real time, and generate a multi-source synchronous dataset of segment units. The sensor acquisition system includes an infrared temperature measurement array and a thermocouple sensor array. The temperature information includes surface temperature and interlayer temperature. The thermal cycling feature extraction module is used to preprocess the temperature information in the segmented unit multi-source synchronous dataset, extract thermal cycling feature parameters, and construct a welding thermal cycling feature parameter set. A welding process constraint generation module is used to obtain planned process parameters, preset line energy and thermal cycle characteristic parameter ranges during the welding process based on material properties and welding process qualification reports, and generate a welding control constraint set. The segmented thermal input deviation calculation module is used to calculate the comprehensive deviation of the thermal input of all welding segment units based on the segmented unit multi-source synchronous dataset, the welding thermal cycle characteristic parameter set and the welding control constraint set, and to obtain the segmented thermal input deviation instruction set. A process parameter adjustment instruction generation module is used to calculate the final current adjustment amount and the final speed adjustment amount using a segmented heat input deviation instruction set, and generate welding process parameter adjustment instructions. The model optimization and update module is used to optimize and adjust the process based on the actual weld microstructure and properties during long-term welding, using parameter optimization algorithms to achieve continuous updating of the segmented heat-controlled welding model.

Images