An automated control system for heater welding assembly

By analyzing real-time welding images and thermal imaging data, the thermal imbalance value is calculated and a thermally symmetrical path is planned. By using the thermal center of gravity to correct deviation, the problem of workpiece deformation caused by the accumulation of thermal shrinkage stress during welding is solved, achieving efficient and precise welding control.

CN123103113APending Publication Date: 2026-11-17FOSHAN YOULE ELECTRIC APPLIANCE CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611455804.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-09-18
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

In existing technologies, welding path planning is guided by the minimum idle stroke, which leads to the accumulation of thermal shrinkage stress in continuous spot welding on one side, forming a bending moment and causing workpiece deformation.

Method used

Welding images and thermal imaging data are acquired through the data acquisition module, the welding status is analyzed, the welding thermal imbalance value is calculated, the welding route is dynamically adjusted, the thermal center of gravity is used as the correction target, the thermal symmetry route is planned to counteract the thermal torque, and welding is prioritized.

Benefits of technology

It effectively corrects the accumulation of thermal shrinkage stress during the welding process, reduces workpiece deformation, improves welding quality and efficiency, and significantly enhances system robustness.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123103113A_ABST
    Figure CN123103113A_ABST
Patent Text Reader

Abstract

This invention discloses an automated control system for heater welding assembly, relating to the field of welding assembly technology. By fusing image and thermal imaging data, the system determines the assembly state through circle fitting and makes decisions based on the welding thermal imbalance value: under normal conditions, welding is performed efficiently along the minimum idle stroke; in abnormal conditions, a thermally symmetrical route is switched to correct the deviation with a reverse thermal torque; after recovery, the system automatically returns to the original path, thereby suppressing thermal stress accumulation and workpiece deformation online. This invention uses real-time thermal imaging to calculate the welding thermal imbalance value to dynamically identify unilateral heat accumulation. In abnormal conditions, the thermal center of gravity of the welded half-circle is obtained using temperature weighting, and welding is prioritized on its directly opposite foot point, using a reverse thermal torque to accurately counteract the bending effect. When there is a conflict with the ideal target, it automatically searches for nearby redundant points to ensure that the welding sequence is always guided by the real-time thermal field. Under normal conditions, it operates efficiently with the minimum idle stroke, only intervening temporarily to correct deviations in cases of thermal imbalance, thus solving the stress deformation problem with minimal cycle time.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of welding assembly technology, specifically to an automated control system for heater welding assembly. Background Technology

[0002] Heater welding assembly is the manufacturing process of connecting heating elements, housings, terminals, and seals into a complete heater assembly through welding. Typically, the metal sheath, flanges, and heating tubes are first tack welded together, and then the main weld seam is completed using methods such as argon arc welding, laser welding, or resistance welding. During the welding process, heat input must be strictly controlled to prevent burning through the thin-walled tube or damaging the internal insulating magnesium oxide powder. After welding, airtightness testing, insulation resistance testing, and withstand voltage testing are required to ensure that the weld seam is leak-free, the electrical insulation is reliable, and the structural strength meets the standards, so that the heater can operate safely and stably for a long time in high-temperature, high-pressure, or corrosive media.

[0003] In the path planning of automated welding, the system usually takes the minimum idle stroke as the optimization goal. For example, it continuously spot welds around the circumference of the heating tube and the flange in the order from 12 o'clock to 1 o'clock to make the welding torch movement trajectory the shortest and the cycle the fastest. However, this pure geometric motion planning completely ignores the dynamic process of thermophysics: when the welding torch continuously ignites the arc in the same half-circle, each weld point will inject extremely high heat into the local base material. The metal in the heated area expands rapidly, but due to the rigid constraint of the surrounding cold metal, it can only produce compressive plastic deformation. When the arc is extinguished and cooled, this area contracts rapidly. Since the spot welding sequence is continuous and sequential, these contraction effects do not exist independently, but accumulate and superimpose along the same side direction to form a continuously increasing bending moment, which eventually pulls the heating tube to the heated side and produces a visible deflection deformation. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides an automated control system for heater welding assembly. This system solves the problem that path planning guided by minimum idle stroke leads to the accumulation of thermal shrinkage stress from continuous spot welding on one side, which in turn generates a bending moment that deforms the workpiece, exposing the fundamental disconnect between pure geometric optimization and thermophysical effects.

[0005] To achieve the above objectives, the present invention provides the following technical solution: an automated control system for heater welding assembly, comprising the following specific modules: a data acquisition module: acquiring welding image data and welding thermal imaging data of welding parts A and B, and performing preprocessing; a welding state analysis module: calculating concentricity and assembly gap based on the welding image data using least-squares circle fitting, and determining the assembly state of A and B to be welded; and a welding route analysis module: calculating the welding thermal imbalance value based on the welding thermal imaging data, analyzing whether there is a thermal imbalance based on the welding thermal imbalance value, and if the analysis indicates normality, continuing welding according to the original preset welding route; if a thermal imbalance is detected, planning a thermally symmetrical route for welding, until normality is achieved, and then resuming welding according to the original preset welding route.

[0006] Furthermore, the calculation method for the welding thermal imbalance value is as follows: when A and B are in the assembly state to be welded, the overlapping part is recorded as the overlapping area, and the overlapping area is divided equally according to the welding thermal imaging data to obtain two areas of equal area. The area welded according to the original welding preset route is recorded as the welding area, and the other area is recorded as the unwelded area. The average temperature values ​​of the welding area and the unwelded area are calculated respectively to obtain the welding temperature value and the unwelded temperature value. The difference between the welding temperature value and the unwelded temperature value is calculated to obtain the welding thermal imbalance value.

[0007] Furthermore, the welding temperature value is obtained by comprehensively calculating the welding temperature value based on the pixel coordinates, number of pixels, and temperature value of each pixel in the welding area.

[0008] Furthermore, the method for analyzing whether there is thermal imbalance based on the welding thermal imbalance value is as follows: a preset thermal imbalance threshold is set, and the welding thermal imbalance value is compared with the thermal imbalance threshold. If the welding thermal imbalance value is greater than the thermal imbalance threshold, it indicates thermal imbalance. If the welding thermal imbalance value is less than or equal to the thermal imbalance threshold, it indicates normal operation.

[0009] Furthermore, the method for planning the thermally symmetrical route for welding is as follows: the center point coordinates are obtained by averaging the number of pixel coordinates and the pixel coordinates. Based on these center point coordinates, the antipodal coordinates corresponding to the set of welded points are calculated using a central symmetry transformation to form a set of candidate target points. Then, the thermal centroid of the welded half-circle is calculated using the current measured temperature as the weight. The antipodal point of this thermal centroid is designated as the primary correction target and welded first. If this thermal centroid conflicts with a welded point, an adjacent redundant point is added, and the sorting continues according to the antipodal principle of the thermal centroid based on the temperature field recovery.

[0010] Furthermore, the method for obtaining the set of soldered joints is as follows: the center coordinates of each solder joint are selected to obtain the set of soldered joints.

[0011] Furthermore, the method for selecting the center coordinates of each solder joint is as follows: using a local temperature maximum detection algorithm, the coordinates of the pixel with the highest temperature value in the heat-affected zone of each solder joint are directly identified as the thermal center coordinates of this solder joint.

[0012] Furthermore, the thermal centroid of the welded half-circle is calculated as follows: the measured temperature value of each pixel in the welding area is used as the weight, and the two-dimensional coordinates are weighted and averaged to obtain the thermal centroid of the welded half-circle.

[0013] Furthermore, the method for determining the direct plantar point as the primary correction target is as follows: the coordinates of the thermal centroid point of the welded half-circle are transformed by 180 degrees of central symmetry, and the resulting direct plantar point is the primary correction target.

[0014] Furthermore, the method for adding the adjacent redundant points is as follows: when the calculated coordinates of the primary correction target point already exist in the set of welded points, the nearest neighbor position that has not been welded and avoids the overlap of the heat-affected zone is searched along the circumferential direction within a preset small angle offset range, with this target point as the center, and it is set as the new primary correction target and added to the welding sequence.

[0015] Compared with the prior art, the embodiments of the present invention have at least the following advantages or beneficial effects: 1. By calculating the welding thermal imbalance value in real time through thermal imaging, abnormal heat accumulation on one side can be dynamically identified, breaking through the limitations of traditional fixed-path pure geometric planning and solving the problem of workpiece bending deformation caused by the cumulative superposition of thermal shrinkage stress.

[0016] 2. By calculating the thermal center of gravity of the welded half-circle with the measured temperature value as the weight, and taking the point directly opposite to it as the primary target for correction and welding, the reverse thermal torque is used to most directly and effectively counteract the maximum bending effect that is accumulating, and the correction action is precise and efficient.

[0017] 3. When the ideal correction target conflicts with the welded joint, the system automatically searches for nearby redundant points within a preset small angle offset range and dynamically sorts them according to the temperature field recovery, ensuring that the welding sequence is always guided by the real-time thermal field rather than a fixed geometric sequence, thus significantly improving the system's robustness.

[0018] 4. Under normal circumstances, the original route with the minimum idle travel is executed to ensure cycle efficiency. Thermal symmetry correction is only temporarily intervened when thermal imbalance is detected. After the situation is restored, the original route is automatically returned. This achieves effective control of residual stress and dimensional accuracy with minimal cycle cost.

[0019] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0020] Figure 1This is a flowchart of the automated control method for heater welding and assembly according to the present invention. Detailed Implementation

[0021] It should be noted that the following detailed description of the technical solutions is based on the accompanying drawings, and is intended to enable those skilled in the art to understand the inventive purpose, technical concept, and beneficial effects of the present invention, and to implement the present invention accordingly. However, these specific embodiments should not be construed as an exhaustive or limiting of the scope of protection of the present invention. Any technical concept disclosed in the present invention, including but not limited to equivalent substitutions of functional modules, reasonable adjustments to the order of steps, or non-substantial modifications to specific parameters in algorithms, as long as they do not depart from the core inventive point of the present invention, namely, the closed-loop control logic of adaptive welding path planning based on real-time thermal field perception using multimodal sensing fusion to actively offset thermal stress deformation, should be considered to fall within the scope of protection sought by this application. Various equivalent implementation methods that can be conceived by those skilled in the art after reading this specification without creative effort also fall within the technical scope covered by the present invention. In other words, the scope of protection of the present invention is determined by the claims, and is not limited to the specific hardware models, software codes, or process parameters exemplified in the specification.

[0022] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus.

[0023] Example 1: like Figure 1 As shown, this embodiment of the invention provides an automated control system for heater welding assembly, comprising the following specific modules: Data acquisition module: The Hikvision MV-CE050-30GM industrial camera acquires welding image data of welded parts A and B. The welding image data includes the number of pixels and the coordinates of the pixels. The Hikvision MV-CE050-30GM industrial camera is a 5-megapixel camera with high resolution, which can provide a sufficient number of pixels and accurate pixel coordinates to ensure the accuracy of subsequent geometric calculations. The welding thermal imaging data of welded parts A and B are acquired by the FLIR A500f thermal imager. The welding thermal imaging data shows the temperature value of each pixel. Welded part A is, for example, an electric heating tube, and welded part B is, for example, a flange. In addition, the FLIR A500f is a thermal imager specifically designed for industrial online monitoring. It can not only output the accurate temperature value of each pixel in the image in real time, but also has high thermal sensitivity and can capture the subtle temperature difference of heat accumulation on one side. Noise reduction processing is performed on both welding image data and welding thermal imaging data. For welding image data, Gaussian filtering is used to suppress impulse noise and random noise caused by arc spatter or smoke during welding, while edge-preserving filtering is used to avoid blurring the key workpiece contour edges. For welding thermal imaging data, since thermal imagers have temporal random fluctuations and spatial non-uniformity, multi-frame temporal mean filtering is first used to stabilize the temperature values ​​of each pixel. Then, bilateral filtering is combined to smooth thermal noise while preserving the temperature gradient boundary of the heat-affected zone of the weld, thereby ensuring the accuracy of subsequent pixel coordinate and temperature value extraction.

[0024] Welding Status Analysis Module: This module performs grayscale processing on welding image data. In this scenario, the color information of the workpiece is not helpful in determining its outline shape; instead, it increases the amount of data and the complexity of the algorithm. Converting the color image to grayscale retains only the brightness information, much like photocopying a color blueprint into a black and white version. This preserves the necessary outline contrast and doubles the running speed of all subsequent image processing algorithms. The resulting pixel grayscale is used for threshold segmentation and edge detection. The Canny edge detection method extracts the outline pixel sets of welded parts A and B. Using the pixel coordinates of these pixels, the least squares circle fitting algorithm is used to accurately calculate the center position and radius of the outer circle of A and the inner hole of B, respectively. Then, the concentricity is obtained based on the difference in the center coordinates, and the uniformity of the assembly gap is calculated based on the radius difference and the pixel distance of the edge point pairs. At the same time, the area is filtered by combining the number of pixels to eliminate interference. When the system determines that the deviation of the center of the two parts is within the preset pixel tolerance and the gap size at all points on the circumference meets the welding process requirements, it is confirmed that A and B have reached the correct assembly state for welding.

[0025] Welding route analysis module: Calculates the welding thermal imbalance value based on welding thermal imaging data. Analyzes whether thermal imbalance exists based on this value. If the analysis indicates normal operation, welding continues according to the original preset welding route. The original preset welding route is the minimum idle distance route. The reason the original preset welding route is usually designed in a circular sequence, such as continuous spot welding from 12 o'clock to 1 o'clock, is because it aims to maximize pure geometric efficiency in engineering. This planning optimizes the welding torch's movement trajectory by minimizing idle distance, maximizing cycle time, while also simplifying the programming logic and eliminating the need for complex coordinate transformations. Frequent large-angle jumps can minimize wear on the robotic arm joints. If thermal imbalance is detected, a thermally symmetrical welding route is planned. In actual welding, due to material gap deviation, uneven heat dissipation of the fixture, or the anisotropic thermal conductivity of the material itself, continuous welding in the original sequence can easily cause heat to accumulate on one side of the circumference, resulting in continuously accumulating shrinkage stress, which eventually bends the heating element. Therefore, the original welding route is the benchmark for efficiency under normal conditions, while the thermally symmetrical route is the correction route for quality in abnormal situations, until the original welding route is restored after normal operation.

[0026] Example 2 differs from Example 1 in that: The calculation method for welding thermal imbalance value is as follows: When A and B are in the assembly state to be welded, the overlapping part is recorded as the overlapping area. Based on the welding thermal imaging data, the overlapping area is divided into two areas of equal area. The area welded according to the original preset welding route is recorded as the welding area, and the other area is recorded as the unwelded area. That is, the system welds along the original minimum clearance route, which itself starts from one side and welds sequentially. Therefore, the most direct heat imbalance is reflected between the semicircle being welded and the semicircle that has not yet been welded. The average temperature values ​​of the welding area and the unwelded area are calculated separately to obtain the welding temperature value and the unwelded temperature value. The difference between the welding temperature value and the unwelded temperature value is calculated to obtain the welding thermal imbalance value. ; in, Indicates the welding thermal imbalance value. This indicates the welding temperature value. This indicates the temperature value before welding. When welding according to the original welding preset route, the welding temperature value is greater than the temperature value before welding. The greater the welding thermal imbalance value, the higher the risk of bending.

[0027] The welding temperature value is obtained as follows: The welding temperature value is obtained by comprehensively calculating the pixel coordinates, number of pixels, and temperature value of each pixel in the welding area. ; in, This indicates the welding temperature value. This indicates the number of pixels in the welding area. Indicates the first in the welding area The temperature values ​​at the coordinates of each pixel are obtained, with the unwelded temperature value being obtained in the same way as the welded temperature value.

[0028] The method for analyzing whether there is thermal imbalance based on the welding thermal imbalance value is as follows: A preset thermal imbalance threshold is established by using finite element simulation software, inputting the workpiece material, wall thickness, and welding process parameters, to simulate the temperature field evolution under the original preset welding sequence, extracting the theoretical critical temperature difference between the welded and unwelded areas, and multiplying it by a safety factor of 0.6 to 0.8 as the initial threshold. Then, offline extreme working condition test welding is performed using standard workpieces, increasing the assembly gap or reducing the fixture pressure, to find the actual temperature difference when irreversible displacement of the pipe end begins, and setting this as the physical intervention threshold. The welding thermal imbalance value is compared with the thermal imbalance threshold. If the welding thermal imbalance value is greater than the thermal imbalance threshold, it indicates thermal imbalance; if the welding thermal imbalance value is less than or equal to the thermal imbalance threshold, it indicates normal operation.

[0029] The method for planning a thermally symmetrical welding route is as follows: The center point coordinates are obtained by averaging the number of pixel coordinates. Based on these center point coordinates, the antipodal coordinates of the welded point set are calculated using a central symmetry transformation, forming a candidate target point set. Then, using the current measured temperature as a weight, the thermal centroid of the welded half-circle is calculated. The antipodal point of this thermal centroid is designated as the primary correction target and welded first, thereby directly and effectively counteracting the accumulating maximum bending effect with the reverse thermal torque. If this thermal centroid conflicts with a welded point, an adjacent redundant point is added, and the welding sequence continues to be sorted according to the antipodal principle of the thermal centroid, depending on the temperature field recovery, ensuring that the welding sequence is always guided by the real-time thermal field rather than a fixed geometric sequence.

[0030] The method for obtaining the set of solder joints is as follows: During welding, each weld point acts as a localized, instantaneous high-temperature heat source. The heat diffuses in a gradient from the center of the weld to the surrounding base material, decreasing with distance. Therefore, by selecting the center coordinates of each weld point, a set of welded points can be obtained. This method does not require additional position sensors or welding torch trajectory recording. It can automatically and accurately identify the position of the welded points directly through the thermal imaging data itself, providing a basis for subsequent antipodal point calculations.

[0031] The method for selecting the center coordinates of each solder joint is as follows: Based on the phenomenon that the heat of the solder joint decreases gradually from the center to the surrounding area, a local temperature maximum detection algorithm is adopted to directly identify the coordinates of the pixel with the highest temperature value in the heat-affected zone of each solder joint as the coordinates of the thermal center of this solder joint.

[0032] The calculation method for the thermal centroid point of the welded halfway point is as follows: Using the measured temperature value of each pixel within the welding area as a weight, a weighted average of its two-dimensional coordinates is performed to obtain the thermal centroid point of the welded half-circle. The higher the temperature of the pixel, the stronger the attraction of the centroid position, thus automatically biasing the calculation results towards the area with the most severe heat accumulation.

[0033] The method of setting the plantar point as the primary correction target is as follows: The coordinates of the thermal centroid point that has been welded for half a cycle are transformed by 180 degrees of central symmetry. The resulting point directly opposite the foot is the first choice for correction. Since it is directly opposite the area where the heat accumulation is most severe, welding here can generate the most direct and effective reverse thermal torque. It is a completely deterministic instantaneous operation with a time complexity of O(1), which ensures that the preceding calculation of the correction path planning will not introduce any delay.

[0034] The method for adding adjacent redundant points is as follows: When the calculated primary correction target point coordinates already exist in the set of welded points, the nearest neighbor position that has not been welded and avoids the overlap of the heat-affected zone is searched along the circumference within a preset small angle offset range, using this target point as the center. That is, when the calculated antipodal point of the thermal center of gravity has been welded, if forced remelting is performed, it will damage the metallurgical integrity of the existing weld and cause the old and new heat-affected zones to overlap, resulting in harmful residual stress superposition or even micro-cracks. Therefore, it is necessary to avoid the welded points and their heat-affected zones, set them as the new primary correction target, and add them to the welding sequence. The preset small angle offset range is ±10 degrees.

[0035] It should be clarified that the specific embodiments described in this specification, particularly the detailed descriptions of industrial camera models, thermal imager models, specific filtering algorithm parameters, pixel tolerance values, and angular offsets, are fundamentally intended to demonstrate to those skilled in the art in a reproducible manner how the technical concept of the present invention can be transformed into a workable physical system, thereby satisfying the patent law's requirement of full disclosure. These specific examples should not be construed as the only path to achieving the technical objectives of the present invention, nor should they be considered as constituting an exclusive limitation on the technical boundaries of the present invention.

[0036] From an abstract technical perspective, the core concept of this invention is not fixed as a specific hardware combination or a particular piece of code. Its essence lies in constructing a closed-loop adaptive welding path planning and control framework based on online thermal field sensing. The core logic chain of this framework includes: using multimodal sensing to acquire welding assembly status and dynamic heat distribution in real time; defining and calculating a quantitative index that characterizes the asymmetry of the thermal field as a criterion for triggering process path switching; executing a preset path with geometric efficiency as the optimization objective under normal conditions, and dynamically switching to a correction path with thermodynamic symmetry as the optimization objective when thermal imbalance is detected; the generation of the correction path is not based on a preset fixed geometric symmetry point, but on real-time temperature field data, locating the center of mass of heat accumulation (i.e., the thermal centroid) through weighted calculation, and welding is performed on its antipodal point to achieve precise physical correction based on the reverse thermal torque; when the ideal correction objective conflicts with physical reality, it has a fault-tolerant mechanism to perform suboptimal search within the constrained neighborhood; finally, after the thermal field returns to equilibrium, it automatically switches back to the efficiency-first normal path to achieve a dynamic balance between optimal efficiency and quality.

[0037] Based on the aforementioned core technology, the technical solution of this invention has the potential to be horizontally transferred and vertically expanded to various industrial welding and additive manufacturing scenarios. Regarding the geometry of the application object, although the embodiment uses a circular circumferential seam between a heating element and a flange as an example, the calculation method for thermal imbalance and the principle of correcting the thermal center of gravity against the foot can be easily extended to welding closed or non-closed paths of any shape. For example, in processes such as four-corner spot welding of rectangular flanges, multi-layer multi-pass welding of long straight welds, or segmented back-welding of large structural components, simply adjusting the area distribution strategy to an appropriate partition that matches the weld direction and heat dissipation boundary conditions, and adjusting the central symmetry transformation to a corresponding spatial symmetry transformation, allows the core idea of ​​this invention to be applied to new scenarios to suppress asymmetric angular deformation or warping caused by continuous welding.

[0038] At the algorithmic level of data acquisition and processing, the noise reduction methods described in this embodiment, such as median filtering, Gaussian filtering, and bilateral filtering, are all general basic operators in the field of signal processing. In actual industrial deployments, more advanced or efficient equivalent algorithms can be flexibly replaced based on the statistical characteristics of on-site noise, the surface emissivity characteristics of the workpiece being inspected, and the computational redundancy of the control system. For example, on embedded platforms with limited computing power, a fast approximation algorithm based on guided filtering can be used for real-time thermal image denoising; for image contour extraction, a lightweight U-Net variant based on deep learning semantic segmentation networks can be used to replace the traditional threshold segmentation and Canny edge detection process, thereby achieving stronger robustness under complex lighting and low-contrast conditions. Any combination of algorithms that can transform sensor data into stable assembly state quantities and thermal field distribution quantities, as long as they serve the same closed-loop decision logic, should be considered to fall within the technical essence scope of this invention.

[0039] At the level of control threshold and decision-making logic, the method for determining the thermal imbalance threshold proposed in the embodiments—a strategy combining finite element simulation and physical calibration under extreme conditions—is itself a universally applicable methodology. The essence of this method is to construct a safety mapping from theoretical critical physical quantities to actual engineering intervention quantities. Under different application objects or process parameters, only the input conditions of the simulation model need to be adjusted and the calibration process re-executed to generate a threshold suitable for the new operating conditions. This threshold generation logic, rather than the threshold value itself, is one of the core components of this invention that ensures the operating condition adaptability of its control strategy.

[0040] Furthermore, the modular architecture design emphasized in this invention endows the system with great flexibility and configurability. The functional division of the data acquisition module, welding status analysis module, and welding path analysis module represents a logical decoupling. In practical distributed control systems, these modules can be deployed on different hardware nodes and interact with each other via industrial Ethernet such as EtherCAT or other real-time buses. For example, image and thermal imaging processing can be completed at the smart camera end, with only the calculated geometric feature values ​​and thermal imbalance values ​​uploaded to the central controller; while path planning and motion control commands are uniformly coordinated by the central controller. Such aggregation, splitting, or cross-hardware deployment of functional modules, as long as the system reconstruction is based on the input-processing-output logical relationship disclosed in this invention, falls within the protection scope of this invention.

[0041] In summary, the text and accompanying drawings in this specification are merely illustrative representations of the principles of this invention, and no specific implementation details constitute a limitation on the claims. The true innovation of this invention lies in providing a novel technological paradigm: through multimodal information fusion, automated welding systems acquire the ability to sense, quantify, and proactively counteract the negative effects of their own thermophysical processes. Any technical solution that achieves the same or similar adaptive suppression of thermal deformation, obtained through reasonable transformation of technical means—including but not limited to: sensor replacement, algorithm upgrades, threshold recalibration, equivalent adjustments to the sequence of steps, and recombination of module functions—under the inspiration of the real-time thermal field-guided path decision-making concept pioneered by this invention, regardless of its application scenario (heater welding, automotive parts manufacturing, or the connection of other precision components), should be understood as falling within the spirit and scope of protection sought by this invention.

Claims

1. An automated control system for heater welding assembly, characterized in that: Includes the following specific modules: Data acquisition module: Acquires welding image data and welding thermal imaging data of weldment A and weldment B, and performs preprocessing; Welding status analysis module: Based on the welding image data, the concentricity and assembly gap are calculated by least squares circle fitting to determine the assembly status of A and B to be welded; Welding route analysis module: Calculates the welding thermal imbalance value based on the welding thermal imaging data, analyzes whether there is a thermal imbalance based on the welding thermal imbalance value. If the analysis shows that it is normal, it continues to weld according to the original preset welding route. If the analysis shows that there is a thermal imbalance, it plans a thermally symmetrical route for welding until it is normal, and then restores the original preset welding route for welding.

2. The automated control system for heater welding assembly according to claim 1, characterized in that: The calculation method for the welding thermal imbalance value is as follows: When A and B are in the assembly state to be welded, the overlapping part is recorded as the overlapping area. The overlapping area is divided into two areas of equal area according to the welding thermal imaging data. The area welded according to the original welding preset route is recorded as the welding area, and the other area is recorded as the unwelded area. The average temperature values ​​of the welding area and the unwelded area are calculated to obtain the welding temperature value and the unwelded temperature value. The difference between the welding temperature value and the unwelded temperature value is calculated to obtain the welding thermal imbalance value.

3. The automated control system for heater welding assembly according to claim 2, characterized in that: The welding temperature value is obtained as follows: The welding temperature value is obtained by comprehensively calculating the pixel coordinates, number of pixels, and temperature value of each pixel in the welding area.

4. The automated control system for heater welding assembly according to claim 3, characterized in that: The method for analyzing whether there is thermal imbalance based on the welding thermal imbalance value is as follows: A preset thermal imbalance threshold is set. The welding thermal imbalance value is compared with the thermal imbalance threshold. If the welding thermal imbalance value is greater than the thermal imbalance threshold, it indicates thermal imbalance. If the welding thermal imbalance value is less than or equal to the thermal imbalance threshold, it indicates normal operation.

5. The automated control system for heater welding assembly according to claim 4, characterized in that: The welding method for the planned thermally symmetrical route is as follows: The center point coordinates are obtained by averaging the number of pixel coordinates and the pixel coordinates. Based on these center point coordinates, the antipodal coordinates of the set of welded points are calculated using a central symmetry transformation, forming a set of candidate target points. Then, the thermal centroid of the welded half-circle is calculated using the current measured temperature as the weight. The antipodal point of this thermal centroid is designated as the primary correction target and welded first. If this thermal centroid conflicts with a welded point, an adjacent redundant point is added, and the sorting continues according to the antipodal principle of the thermal centroid based on the temperature field recovery.

6. The automated control system for heater welding assembly according to claim 5, characterized in that: The method for obtaining the set of solder joints is as follows: The center coordinates of each solder joint are selected to obtain the set of solder joints.

7. The automated control system for heater welding assembly according to claim 6, characterized in that: The method for selecting the center coordinates of each solder joint is as follows: A local temperature maximum detection algorithm is used to directly identify the coordinates of the pixel with the highest temperature value in the heat-affected zone of each solder joint as the thermal center coordinates of that solder joint.

8. The automated control system for heater welding assembly according to claim 7, characterized in that: The calculation method for the thermal centroid point of the welded half-circle is as follows: Using the measured temperature value of each pixel within the welding area as a weight, a weighted average of its two-dimensional coordinates is performed to obtain the thermal centroid point of the half-circle welded.

9. The automated control system for heater welding assembly according to claim 8, characterized in that: The method of determining the plantar point as the primary correction target is as follows: The coordinates of the thermal centroid point that has been welded for half a cycle are transformed by 180 degrees of central symmetry, and the resulting point directly opposite the plantar point is the primary target for correction.

10. The automated control system for heater welding assembly according to claim 9, characterized in that: The method for adding adjacent redundant points is as follows: When the calculated primary correction target point coordinates already exist in the set of welded points, the nearest neighbor position that has not been welded and avoids overlapping heat-affected zones is searched along the circumference within a preset small angular offset range, with this target point as the center. This nearest neighbor position is set as the new primary correction target and added to the welding sequence.