Multi-station collaborative automobile part welding device and method

Abstract

The invention discloses a multi-station collaborative automobile part welding device and method, and relates to the technical field of intelligent welding and manufacturing of automobile parts, at least two six-axis industrial robots are distributed to execute a main welding task and an auxiliary stress regulation and control task respectively, and welding deformation is dynamically counteracted through interference of stress waves generated by two heat sources; stress wave phase information is collected in real time through an on-line monitoring device, the output phase of the auxiliary heat source is dynamically adjusted through the phase-locked loop technology, and it is ensured that stress waves of the two heat sources always keep anti-phase interference. According to the method, the thermal sensitivity digital twinborn model is constructed, the coupling layout of the main heat source and the auxiliary heat source is achieved in combination with the optimization algorithm, welding thermal deformation is actively counteracted through stress wave interference, and the welding size precision of automobile parts is greatly improved. Real-time synchronization of heat source phases is achieved through the phase-locked loop technology, and it is ensured that the deformation counteracting effect is stable and reliable. The roles of the main robot and the auxiliary robot are dynamically switched, the effectiveness of deformation regulation and control in the welding process is guaranteed, and residual stress is reduced.

Description

Technical Field

[0001] This invention relates to the field of intelligent welding manufacturing technology for automotive parts, and in particular to a multi-station collaborative welding device and method for automotive parts. Background Technology

[0002] Thermal deformation is a common technical challenge in the welding process of automotive parts, especially for large, thin-walled components, where it directly affects dimensional accuracy and assembly performance. In traditional multi-station welding processes, thermal deformation is considered an inevitable byproduct of the welding process, typically addressed passively by post-weld straightening or increasing fixture rigidity. This passive approach not only fails to eliminate deformation but also easily leads to residual stress concentration within the workpiece, reducing the structural strength and service life of the components. Existing multi-station welding technologies fail to achieve proactive control over welding thermal deformation, making it difficult to meet the high-precision, high-reliability manufacturing requirements of automotive parts, and welding efficiency and process stability need further improvement. Summary of the Invention

[0003] To address the aforementioned technical problems, this invention provides a multi-station collaborative automotive parts welding apparatus and method. The technical solution adopted is as follows:

[0004] A multi-station collaborative welding method for automotive parts includes the following steps:

[0005] Step 1: Construct a digital twin model of the thermal sensitivity of the welded workpiece to obtain the correlation between the workpiece's thermal input and deformation response;

[0006] Step 2: Based on the thermal sensitivity digital twin model, the auxiliary heat source layout scheme is solved by optimization algorithm to form a coupled layout of the main welding heat source and the auxiliary stress control heat source, so that the auxiliary heat source generates a counteracting stress field with the opposite phase to the welding thermal deformation stress.

[0007] Step 3: Assign at least two six-axis industrial robots to perform the main welding task and the auxiliary stress control task respectively. The welding deformation is dynamically offset by the stress wave interference generated by the two heat sources.

[0008] Step 4: Real-time acquisition of stress wave phase information through online monitoring device, and dynamic adjustment of auxiliary heat source output phase using phase-locked loop technology to ensure that the stress waves of the two heat sources always maintain anti-phase interference;

[0009] Step 5: Calculate the workpiece thermal equilibrium point position in real time, and dynamically switch the roles of the main and auxiliary six-axis industrial robots according to the thermal equilibrium point migration to ensure that the welding task is performed by the heat source that is most effective in controlling deformation.

[0010] Optionally, in step 1, the digital twin model of thermal sensitivity is constructed through finite element modal analysis, with the core parameter being the thermal sensitivity matrix. , defined as any discrete point on the workpiece surface Under a unit heat input, the displacement response matrix of each node of the workpiece is expressed as:

[0011] ;

[0012] in, This represents the total number of discrete units on the workpiece surface. This represents the total number of monitoring nodes for the entire workpiece. For the first The coordinates are When a unit heat pulse is applied to the discrete unit, the first The displacement response generated by each monitoring node.

[0013] Optionally, the finite element modal analysis method is as follows: using finite element simulation software, the workpiece is discretized into several tetrahedral elements, with the element size set to 0.5mm-2mm according to the workpiece wall thickness; heat pulses with an amplitude of 0.8W-1W and a duration of 0.1-0.15s are applied to each discrete element; through transient thermal analysis and structural mechanical analysis, the displacement response of each monitoring node of the entire workpiece within 0-10s after the heat input is recorded; and a thermal sensitivity matrix is ​​constructed based on the response data. The monitoring nodes are evenly distributed on the surface of the workpiece, with a node spacing of no more than 5 mm.

[0014] Optionally, the optimization algorithm in step 2 is a genetic algorithm, with the objective function being minimizing post-weld residual deformation, and the location, power, and heating sequence of the auxiliary heat source being used as variables for iterative optimization. The objective function expression is: ;

[0015] in, This represents the total residual deformation after welding. The predicted deformation amount generated by the main welding heat source at the j-th monitoring node. To compensate for the deformation generated by the heat source at the j-th monitoring node, a thermodynamic sensitivity matrix is ​​introduced as a weighting factor in the fitness function.

[0016] Optionally, the iterative optimization process of the genetic algorithm is as follows: the population size is initialized to 50-100, the search range of the heat source location is the welding area of ​​the workpiece and within 50mm of its perimeter, the power search range is 0.3 times the main welding power to the main welding power, the heating time sequence search step size is 0.1s; the number of iterations is set to 30-50 generations, the crossover probability is 0.6-0.8, the mutation probability is 0.05-0.1, and the convergence of the objective function is used as the iteration termination condition.

[0017] Optionally, the stress wave interference in step 3 can be implemented as follows: the main welding heat source operates at a constant power. Welding, with a frequency of Thermal stress waves; auxiliary heat source with power TIG remelting scanning is performed in a specific area of ​​the workpiece, and the movement speed of the auxiliary heat source or the frequency of the arc pulse is controlled to adjust the frequency of the generated thermal stress wave. satisfy ,in It is a frequency fine-tuning amount of 0.1-1Hz, used to adapt to dynamic phase drift.

[0018] Optional TIG remelting scanning parameters for the auxiliary heat source are: scanning speed 5mm / s-15mm / s, arc voltage 10V-15V, and argon protective gas flow rate 8L / min-12L / min; under arc heating mode, the arc pulse frequency and thermal stress wave frequency are... Consistent, with a pulse duty cycle of 30%-50%.

[0019] Optionally, the online monitoring device in step 4 is a laser Doppler vibrometer array, used to calculate the instantaneous phase of the stress wave on the workpiece surface in real time. Phase-locked loop (PLL) technology uses a closed-loop system consisting of phase detection, loop filtering, and voltage-controlled oscillation to adjust the inverter frequency of the auxiliary heat source power supply, forcing the auxiliary heat source to output a different phase. satisfy To achieve anti-phase synchronization;

[0020] The method for phase-locked loops is as follows: calculate the measured phase. The difference between the phase and the theoretical output Using a PI controller Filtering generates the control voltage U(t), with a PI controller proportional coefficient of 0.5-2.0 and an integral coefficient of 0.1-0.5; the inverter frequency of the auxiliary heat source power supply is adjusted via a numerically controlled oscillator until... Phase locking is achieved.

[0021] Optionally, the thermal equilibrium point in step 5 is the instantaneous neutral surface B(t) where the internal stress of the workpiece is zero, i.e. The contour lines are obtained; the local elastic modulus of the workpiece is updated based on the current temperature field, the thermal sensitivity matrix is ​​recalculated, and efficiency is adjusted. , ;in The displacement of the thermal equilibrium point caused by a unit of heat. Heat is input to the heat source; when the original welding six-axis industrial robot's control efficiency is lower than the threshold of 0.3mm / kJ-0.5mm / kJ, and another six-axis industrial robot is higher than this threshold, a role swap is triggered.

[0022] A multi-station collaborative automotive parts welding device is disclosed to implement a multi-station collaborative automotive parts welding method. The device includes at least two six-axis industrial robots, welding torches matching the number of six-axis industrial robots, a welding controller, a laser Doppler vibration meter array, a flexible tooling platform, and a three-dimensional structured light scanner. The welding torches are detachably mounted on the movable ends of the six-axis industrial robots. The welding controller includes a thermo-coupling simulation module, a genetic algorithm optimization module, and a digital phase-locked loop (PLL) controller. The thermo-coupling simulation module is used to construct a thermo-sensitive digital twin model and calculate and predict the deformation field. The genetic algorithm optimization module is used to solve for the auxiliary heat source layout scheme. The PLL controller is used to achieve real-time synchronization of the auxiliary heat source phase. The laser Doppler vibration meter array is arranged in a mesh above the welding station to non-contactly collect the stress wave propagation velocity, phase, and amplitude information of the workpiece surface in real time. The flexible tooling platform is used to support the workpiece. The flexible tooling platform includes a support base plate and several sets of universal adaptive support units. The universal adaptive support unit includes a ball head support rod and an electromagnetic lock. During welding, the electromagnetic lock is released, causing the ball head support rod to displace with the thermal deformation of the workpiece. The three-dimensional structured light scanner is installed on the top of the flexible tooling platform to acquire the three-dimensional point cloud data of the workpiece after welding for residual deformation detection and feedback optimization. The welding controller is communicatively connected to the six-axis industrial robot, welding torch, laser Doppler vibration meter array, and three-dimensional structured light scanner.

[0023] In summary, the present invention has at least one of the following beneficial technical effects:

[0024] This invention provides a multi-station collaborative welding device and method for automotive parts. By constructing a thermally sensitive digital twin model and combining it with optimization algorithms to achieve a coupled layout of main and auxiliary heat sources, stress wave interference is used to actively counteract welding thermal deformation, significantly improving the dimensional accuracy of automotive parts welding. Phase-locked loop (PLL) technology enables real-time synchronization of the heat source phases, ensuring stable and reliable deformation counteracting. Dynamic switching between main and auxiliary robot roles ensures effective deformation control during welding, reducing residual stress generation. The entire method and device achieve active control of welding deformation, reducing post-weld straightening processes, improving welding efficiency and part qualification rate, lowering manufacturing costs, and adapting to the needs of large-scale, high-precision automotive parts manufacturing. Attached Figure Description

[0025] Figure 1 This is a schematic diagram of the structure of a multi-station collaborative automotive parts welding device according to the present invention;

[0026] Figure 2 This is a schematic diagram of the combined structure of a six-axis industrial robot and welding torch for a multi-station collaborative automotive parts welding device according to the present invention.

[0027] Figure 3This is a schematic diagram of the electrical component connection principle of a multi-station collaborative automotive parts welding device according to the present invention.

[0028] Figure reference numerals: 1. Six-axis industrial robot; 2. Welding torch; 3. Welding controller; 31. Thermo-coupling simulation module; 32. Genetic algorithm optimization module; 33. Digital phase-locked loop controller; 4. Laser Doppler vibration meter array; 5. Image processing unit; 6. Flexible tooling platform; 7. Support base plate; 8. Universal adaptive support unit; 9. Ball head support rod; 10. Electromagnetic lock; 11. 3D structured light scanner. Detailed Implementation

[0029] The present invention will be further described in detail below with reference to the accompanying drawings.

[0030] This invention discloses a multi-station collaborative automotive parts welding device and method.

[0031] Reference Figures 1-3 Example 1: A multi-station collaborative welding method for automotive parts, comprising the following steps:

[0032] Step 1: Construct a digital twin model of the thermal sensitivity of the welded workpiece to obtain the correlation between the workpiece's thermal input and deformation response;

[0033] Step 2: Based on the thermal sensitivity digital twin model, the auxiliary heat source layout scheme is solved by optimization algorithm to form a coupled layout of the main welding heat source and the auxiliary stress control heat source, so that the auxiliary heat source generates a counteracting stress field with the opposite phase to the welding thermal deformation stress.

[0034] Step 3: Assign at least two six-axis industrial robots 1 to perform the main welding task and the auxiliary stress control task respectively, and dynamically offset the welding deformation through the stress wave interference generated by the two heat sources;

[0035] Step 4: Real-time acquisition of stress wave phase information through online monitoring device, and dynamic adjustment of auxiliary heat source output phase using phase-locked loop technology to ensure that the stress waves of the two heat sources always maintain anti-phase interference;

[0036] Step 5: Calculate the workpiece thermal equilibrium point position in real time, and dynamically switch the roles of the main and auxiliary six-axis industrial robots according to the thermal equilibrium point migration to ensure that the welding task is performed by the heat source that is most effective in controlling deformation.

[0037] By adopting the above technical solution, Step 1 constructs a digital twin model of thermal sensitivity, establishing the correlation between workpiece thermal input and deformation response, providing a data foundation for subsequent deformation control. Step 2, based on this model, determines the coupling layout of the main welding heat source and the auxiliary stress control heat source through an optimized algorithm, ensuring that the offsetting stress field generated by the auxiliary heat source is out of phase with the welding thermal deformation stress, providing a prerequisite for deformation offsetting. Step 3 assigns at least two six-axis industrial robots 1 to perform the main welding and auxiliary stress control tasks respectively, utilizing the stress wave interference effect generated by the two heat sources to dynamically offset the thermal deformation generated during welding. Step 4 collects stress wave phase information in real time through an online monitoring device, and dynamically adjusts the output phase of the auxiliary heat source using phase-locked loop technology to ensure that the stress waves of the two heat sources always maintain anti-phase interference, ensuring stable deformation offsetting effect. Step 5 calculates the workpiece thermal equilibrium point position in real time, and dynamically switches the roles of the main and auxiliary six-axis industrial robots 1 according to the thermal equilibrium point migration, ensuring that the welding task is always performed by the heat source most effective for deformation control, further improving deformation control accuracy and welding reliability.

[0038] In Example 2, the digital twin model of thermal sensitivity in step 1 is constructed through finite element modal analysis, with the core parameter being the thermal sensitivity matrix. , defined as any discrete point on the workpiece surface Under a unit heat input, the displacement response matrix of each node of the workpiece is expressed as:

[0039] ;

[0040] in, This represents the total number of discrete units on the workpiece surface. This represents the total number of monitoring nodes for the entire workpiece. For the first The coordinates are When a unit heat pulse is applied to the discrete unit, the first The displacement response generated by each monitoring node.

[0041] By adopting the above technical solution and constructing the model using the finite element modal analysis method, the core parameter of which is the thermal sensitivity matrix. Thermal sensitivity matrix Used to characterize arbitrary discrete points on the workpiece surface Under a unit heat input, the displacement response law of each node of the workpiece is induced, and its expression is: Where m represents the total number of discrete units on the workpiece surface, and n represents the total number of monitoring nodes on the entire workpiece. The matrix represents the displacement response of the j-th monitoring node when a unit heat pulse is applied to the i-th discrete element with coordinates x and y. This matrix can accurately quantify the correlation between heat input and overall deformation in each region of the workpiece, providing precise data support for subsequent optimization algorithms and deformation control.

[0042] Example 3: The finite element modal analysis method is as follows: Using finite element simulation software, the workpiece is discretized into several tetrahedral elements, with the element size set to 0.5mm-2mm according to the workpiece wall thickness; a heat pulse with an amplitude of 0.8W-1W and a duration of 0.1-0.15s is applied to each discrete element; through transient thermal analysis and structural mechanical analysis, the displacement response of each monitoring node of the entire workpiece within 0-10s after the heat input is recorded; and a thermal sensitivity matrix is ​​constructed based on the response data. The monitoring nodes are evenly distributed on the surface of the workpiece, with a node spacing of no more than 5 mm.

[0043] By adopting the above technical solution and using finite element simulation software for analysis, the workpiece is first discretized into several tetrahedral elements. The element size is set to 0.5mm-2mm according to the workpiece wall thickness to ensure that the simulation accuracy matches the actual situation of the workpiece. Then, heat pulses with an amplitude of 0.8W-1W and a duration of 0.1-0.15s are applied to each discrete element. The temperature change law of each region of the workpiece is obtained through transient thermal analysis. Combined with structural mechanics analysis, the deformation response of the workpiece is obtained. The displacement response data of each monitoring node of the entire workpiece within 0-10s after the heat input are recorded. The monitoring nodes are evenly distributed on the surface of the workpiece, with a node spacing of no more than 5mm to ensure the comprehensiveness and accuracy of the displacement response data. Based on this data, a thermal sensitivity matrix is ​​constructed. This ensures that the matrix accurately reflects the relationship between the workpiece's thermal input and deformation response.

[0044] Example 4: The optimization algorithm in step 2 is a genetic algorithm, with minimizing post-weld residual deformation as the objective function, and iterative optimization using the location, power, and heating sequence of the auxiliary heat source as variables. The objective function expression is: ;

[0045] in, This represents the total residual deformation after welding. The predicted deformation amount generated by the main welding heat source at the j-th monitoring node. To compensate for the deformation generated by the heat source at the j-th monitoring node, a thermodynamic sensitivity matrix is ​​introduced as a weighting factor in the fitness function.

[0046] By adopting the above technical solution, the optimization algorithm uses a genetic algorithm, with minimizing post-weld residual deformation as the core objective, and constructs the objective function. Where ∆D represents the total residual deformation after welding. The predicted deformation amount generated by the main welding heat source at the j-th monitoring node. The auxiliary heat source is used to offset the deformation generated at the j-th monitoring node. The genetic algorithm uses the location, power, and heating sequence of the auxiliary heat source as iterative variables, and introduces the thermal sensitivity matrix as a weighting factor in the fitness function. It prioritizes the placement of auxiliary heat sources in regions with high deformation control efficiency, and obtains the optimal auxiliary heat source layout scheme through iterative optimization. This ensures that the auxiliary heat source can generate a offsetting stress field with the opposite phase to the welding thermal deformation stress, thereby achieving precise offsetting of welding deformation.

[0047] Example 5: The iterative optimization process of the genetic algorithm is as follows: the population size is initialized to 50-100, the search range of the heat source location is the welding area of ​​the workpiece and within 50mm of its perimeter, the power search range is 0.3 times the main welding power to the main welding power, the heating time sequence search step size is 0.1s, the number of iterations is set to 30-50 generations, the crossover probability is 0.6-0.8, the mutation probability is 0.05-0.1, and the convergence of the objective function is used as the iteration termination condition.

[0048] By adopting the above technical solution and reasonably setting various parameters of the genetic algorithm, the optimization efficiency and accuracy are ensured. The initial population size is set to 50-100 to ensure population diversity while avoiding excessive computation. The search range for the heat source location is limited to the workpiece welding area and its surrounding 50mm. The power search range is set to 0.3 times the main welding power to the main welding power. The heating sequence search step size is set to 0.1s to ensure that the search range and step size match the actual welding scenario. The number of iterations is set to 30-50 generations, the crossover probability is set to 0.6-0.8, and the mutation probability is set to 0.05-0.1 to avoid the algorithm getting trapped in local optima while ensuring the algorithm's convergence speed. The convergence of the objective function is used as the iteration termination condition, and the optimal auxiliary heat source layout scheme is finally output.

[0049] Example 6, the specific implementation of stress wave interference in step 3 is as follows: the main welding heat source operates at a constant power. Welding, with a frequency of Thermal stress waves; auxiliary heat source with power TIG remelting scanning is performed in a specific area of ​​the workpiece, and the movement speed of the auxiliary heat source or the frequency of the arc pulse is controlled to adjust the frequency of the generated thermal stress wave. satisfy ,in It is a frequency fine-tuning amount of 0.1-1Hz, used to adapt to dynamic phase drift.

[0050] By adopting the above technical solution, the main welding heat source operates at a constant power. Welding is performed, and the frequency generated during the welding process is... The thermal stress wave causes thermal deformation of the workpiece. An auxiliary heat source with power... TIG remelting scanning is performed in a specific area of ​​the workpiece. By precisely controlling the movement speed of the auxiliary heat source or the frequency of the arc pulse, the frequency of the thermal stress wave generated by the auxiliary heat source is adjusted. satisfy ∆f is a frequency fine-tuning amount of 0.1-1Hz, used to adapt to the dynamic drift of the stress wave phase during welding, ensuring that the stress waves generated by the two heat sources can form a stable anti-phase interference, and cancel each other out through the interference effect, thereby achieving dynamic suppression of welding deformation.

[0051] Example 7: The TIG remelting scanning parameters for the auxiliary heat source were: scanning speed 5mm / s-15mm / s, arc voltage 10V-15V, and argon protective gas flow rate 8L / min-12L / min; under arc heating mode, the arc pulse frequency and thermal stress wave frequency were... Consistent, with a pulse duty cycle of 30%-50%.

[0052] By adopting the above technical solution and setting reasonable TIG remelting scanning parameters, it is ensured that the auxiliary heat source can generate a stable stress wave to counteract the stress. The scanning speed is set to 5mm / s-15mm / s, the arc voltage is set to 10V-15V, and the argon shielding gas flow rate is set to 8L / min-12L / min. This parameter range can ensure the stability of the TIG remelting scan and avoid insufficient or excessive remelting. In the arc heating mode, the arc pulse frequency is kept consistent with the thermal stress wave frequency f2, and the pulse duty cycle is set to 30%-50%. This ensures that the amplitude of the thermal stress wave generated by the auxiliary heat source is stable and forms effective anti-phase interference with the thermal stress wave generated by the main welding heat source, thereby improving the deformation counteracting effect.

[0053] In Example 8, the online monitoring device in step 4 is a laser Doppler vibrometer array, used to calculate the instantaneous phase of the stress wave on the workpiece surface in real time. Phase-locked loop (PLL) technology uses a closed-loop system consisting of phase detection, loop filtering, and voltage-controlled oscillation to adjust the inverter frequency of the auxiliary heat source power supply, forcing the auxiliary heat source to output a different phase. satisfy To achieve anti-phase synchronization;

[0054] The method for phase-locked loops is as follows: calculate the measured phase. The difference between the phase and the theoretical output Using a PI controller Filtering generates the control voltage U(t), with a PI controller proportional coefficient of 0.5-2.0 and an integral coefficient of 0.1-0.5; the inverter frequency of the auxiliary heat source power supply is adjusted via a numerically controlled oscillator until... Phase locking is achieved.

[0055] By adopting the above technical solution, the phase-locked loop (PLL) technology achieves real-time synchronization of the auxiliary heat source phase through a closed-loop system composed of phase detection, loop filtering, and voltage-controlled oscillation. First, the difference between the measured phase θ1t and the theoretical output phase is calculated. Using a PI controller Filtering is performed to eliminate high-frequency noise and generate a stable control voltage U(t). The PI controller's proportional coefficient is set to 0.5-2.0, and the integral coefficient is set to 0.1-0.5 to ensure filtering effectiveness and control response speed. The control voltage U(t) is input to the numerically controlled oscillator, and the inverter frequency of the auxiliary heat source power supply is adjusted until... Phase locking is achieved to ensure the output phase of the auxiliary heat source. satisfy This ensures that the stress waves from the two heat sources always remain in opposite phases and interfere.

[0056] In Example 9, the thermal equilibrium point in step 5 is the instantaneous neutral surface B(t) where the internal stress of the workpiece is zero, i.e. The contour lines are obtained; the local elastic modulus of the workpiece is updated based on the current temperature field, the thermal sensitivity matrix is ​​recalculated, and efficiency is adjusted. , ;in The displacement of the thermal equilibrium point caused by a unit of heat. Heat is input to the heat source; when the original welding six-axis industrial robot 1's control efficiency is lower than the threshold of 0.3mm / kJ-0.5mm / kJ, and another six-axis industrial robot 1 is higher than this threshold, the role swap is triggered.

[0057] By adopting the above technical solution, a six-axis industrial robot 1 serves as the actuator, and a welding torch 2 is detachably mounted on the movable end of the six-axis industrial robot 1 to perform main welding and auxiliary stress control tasks. A welding controller 3 serves as the control core; its internal thermo-coupling simulation module 31 is used to construct a thermo-sensitive digital twin model and calculate and predict the deformation field, a genetic algorithm optimization module 32 is used to solve the auxiliary heat source layout scheme, and a digital phase-locked loop controller is used to achieve real-time synchronization of the auxiliary heat source phase. A laser Doppler vibration meter array 4 is arranged in a mesh above the welding station, non-contactly acquiring the propagation speed, phase, and amplitude information of stress waves on the workpiece surface in real time, providing data support for phase-locked loop technology. A flexible tooling platform 5 supports the workpiece, including a support base plate 51 and a universal adaptive support unit. During welding, the electromagnetic lock 522 is released, causing the ball-head support rod 521 to displace with the thermal deformation of the workpiece, avoiding the influence of rigid constraints on thermal stress release. A three-dimensional structured light scanner 6 is installed on top of the flexible tooling platform 5 to acquire three-dimensional point cloud data of the workpiece after welding, providing data for residual deformation detection and process feedback optimization. The welding controller 3 is connected to the six-axis industrial robot 1, the welding torch 2, the laser Doppler vibration meter array 4, and the three-dimensional structured light scanner 6 to achieve coordinated control of the components and ensure the stable operation of the entire device.

[0058] Example 10: A multi-station collaborative automotive parts welding device for implementing a multi-station collaborative automotive parts welding method. The device includes at least two six-axis industrial robots 1, welding torches 2 matching the number of six-axis industrial robots 1, a welding controller 3, a laser Doppler vibration meter array 4, a flexible tooling platform 5, and a three-dimensional structured light scanner 6. The welding torches 2 are detachably mounted on the movable end of the six-axis industrial robots 1. The welding controller 3 includes a thermo-coupling simulation module 31, a genetic algorithm optimization module 32, and a digital phase-locked loop controller 33. The thermo-coupling simulation module 31 is used to construct a thermo-sensitive digital twin model and calculate and predict the deformation field; the genetic algorithm optimization module 32 is used to solve the auxiliary heat source layout scheme; the digital phase-locked loop controller is used to realize real-time synchronization of the auxiliary heat source phase. The laser Doppler vibration meter array 4 is arranged in a mesh above the welding station to non-contactly collect the stress wave propagation speed, phase, and amplitude information of the workpiece surface in real time. The flexible tooling platform 5 is used to support the workpiece. The flexible tooling platform 5 includes a support base plate 51 and several sets of universal adaptive support units. The universal adaptive support unit includes a ball head support rod 521 and an electromagnetic locker 522. During the welding process, the electromagnetic locker 522 is released, causing the ball head support rod 521 to move with the thermal deformation of the workpiece. The three-dimensional structured light scanner 6 is installed on the top of the flexible tooling platform 5 to acquire the three-dimensional point cloud data of the workpiece after welding for residual deformation detection and feedback optimization. The welding controller 3 is communicatively connected to the six-axis industrial robot 1, the welding torch 2, the laser Doppler vibration meter array 4, and the three-dimensional structured light scanner 6.

[0059] By adopting the above technical solution, a six-axis industrial robot 1 serves as the actuator, and a welding torch 2 is detachably mounted on the movable end of the six-axis industrial robot 1 to perform main welding and auxiliary stress control tasks. A welding controller 3 serves as the control core; its internal thermo-coupling simulation module 31 is used to construct a thermo-sensitive digital twin model and calculate and predict the deformation field, a genetic algorithm optimization module 32 is used to solve the auxiliary heat source layout scheme, and a digital phase-locked loop controller is used to achieve real-time synchronization of the auxiliary heat source phase. A laser Doppler vibration meter array 4 is arranged in a mesh above the welding station, non-contactly acquiring the propagation speed, phase, and amplitude information of stress waves on the workpiece surface in real time, providing data support for phase-locked loop technology. A flexible tooling platform 5 supports the workpiece, including a support base plate 51 and a universal adaptive support unit. During welding, the electromagnetic lock 522 is released, causing the ball-head support rod 521 to displace with the thermal deformation of the workpiece, avoiding the influence of rigid constraints on thermal stress release. A three-dimensional structured light scanner 6 is installed on top of the flexible tooling platform 5 to acquire three-dimensional point cloud data of the workpiece after welding, providing data for residual deformation detection and process feedback optimization. The welding controller 3 is connected to the six-axis industrial robot 1, the welding torch 2, the laser Doppler vibration meter array 4, and the three-dimensional structured light scanner 6 to achieve coordinated control of the components and ensure the stable operation of the entire device.

[0060] The following specific embodiments illustrate the implementation principle of the present invention:

[0061] Taking the welding of automotive frame crossbeams as a specific application example, the multi-station collaborative automotive component welding method and apparatus of the present invention are used to achieve high-precision welding of frame crossbeams. The specific process is as follows.

[0062] The device is a multi-station collaborative automotive parts welding unit, comprising two six-axis industrial robots 1, two welding torches 2 matched with the six-axis industrial robots 1, a welding controller 3, a laser Doppler vibration meter array 4, a flexible tooling platform 5, and a 3D structured light scanner 6. The welding torches 2 are detachably mounted on the movable end of the six-axis industrial robots 1 and can be switched to TIG remelting torches. The welding controller 3 includes a thermo-coupling simulation module 31, a genetic algorithm optimization module 32, and a digital phase-locked loop controller 33. The laser Doppler vibration meter array 4 is arranged in a mesh above the welding stations. The flexible tooling platform 5 includes a support base plate 51 and several sets of omnidirectional adaptive support units, each including a ball-head support rod 521, an electromagnetic lock 522, and a temperature sensor 523. The 3D structured light scanner 6 is mounted on top of the flexible tooling platform 5. The welding controller 3 is communicatively connected to the six-axis industrial robots 1, the welding torches 2, the laser Doppler vibration meter array 4, and the 3D structured light scanner 6.

[0063] The specific steps for welding automotive frame crossbeams using the aforementioned device are as follows:

[0064] Step 1: Place the automobile frame crossbeam workpiece on the flexible tooling platform 5. Using the thermo-mechanical coupling simulation module 31 of the welding controller 3, construct a digital twin model of the thermal sensitivity of the frame crossbeam to obtain the correlation between the thermal input and deformation response of the frame crossbeam.

[0065] Step 2: The genetic algorithm optimization module 32 of the welding controller 3 is based on the thermal sensitivity digital twin model constructed in Step 1. It uses a genetic algorithm to solve the auxiliary heat source layout scheme, forming a coupled layout of the main welding heat source and the auxiliary stress control heat source, so that the auxiliary heat source generates a counteracting stress field that is opposite in phase to the welding thermal deformation stress.

[0066] Step 3: Two six-axis industrial robots 1 are assigned to perform the main welding task and the auxiliary stress control task, respectively. One six-axis industrial robot 1 carries a welding torch 2 as the main welding heat source, while the other six-axis industrial robot 1 carries a TIG remelting torch as the auxiliary stress control heat source. The stress wave interference generated by the two heat sources dynamically counteracts the thermal deformation generated during the welding of the frame crossbeam. The main welding heat source operates at a constant power. Welding, with a frequency of Thermal stress waves; auxiliary heat source with power TIG remelting scanning was performed in a specific area of ​​the frame crossbeam, and the movement speed of the auxiliary heat source was controlled to adjust the frequency of the generated thermal stress wave. satisfy ∆f was set to 0.5Hz. The TIG remelting scan parameters for the auxiliary heat source were set as follows: scan speed 10mm / s, arc voltage 12V, argon protective gas flow rate 10L / min, and arc pulse frequency and thermal stress wave frequency... Consistent, the pulse duty cycle is set to 40%.

[0067] Step 4: The laser Doppler vibration meter array 4 acquires the stress wave phase information on the surface of the vehicle frame crossbeam in real time, and calculates the instantaneous phase of the stress wave in real time. The data is then transmitted to the welding controller 3. The digital phase-locked loop controller 33 of the welding controller 3 utilizes phase-locked loop technology, employing a closed-loop system consisting of phase detection, loop filtering, and voltage-controlled oscillation, to dynamically adjust the output phase of the auxiliary heat source. A PI controller is used to monitor the measured phase. The difference between the phase and the theoretical output After filtering, the PI controller's proportional coefficient is set to 1.2 and its integral coefficient to 0.3, generating a stable control voltage U(t). The inverter frequency of the auxiliary heat source power supply is adjusted via a numerically controlled oscillator until... ≤0.01 Phase locking is achieved to ensure the output phase of the auxiliary heat source. satisfy = This ensures that the stress waves from the two heat sources always remain in opposite phases and interfere.

[0068] Step 5: The welding controller 3 calculates the location of the thermal equilibrium point inside the frame crossbeam in real time. The thermal equilibrium point is the instantaneous neutral surface B(t) where the stress inside the frame crossbeam is zero, i.e. The contour lines. Temperature sensor 523 collects temperature data at the contact points of the frame crossbeam in real time and transmits it to welding controller 3. Welding controller 3 updates the local elastic modulus of the frame crossbeam based on the current temperature field, recalculates the thermal sensitivity matrix, and determines the roles of the two six-axis industrial robots 1 by adjusting the efficiency η. The calculation method of adjustment efficiency η is as follows: When the control efficiency of the original six-axis industrial robot 1 is lower than the threshold of 0.4 mm / kJ, and the control efficiency of another six-axis industrial robot 1 is higher than the threshold, a role swap is triggered to ensure that the welding task is always performed by the six-axis industrial robot 1 that is most effective in controlling deformation.

[0069] After welding, the 3D structured light scanner 6 acquires the 3D point cloud data of the frame crossbeam after welding, which is used for residual deformation detection and feedback optimization to ensure that the welding dimensional accuracy of the frame crossbeam meets the requirements of automobile assembly.

[0070] In this embodiment, finite element modal analysis is performed using finite element simulation software. The frame crossbeam is discretized into several tetrahedral elements, with the element size set to 1 mm based on the frame crossbeam wall thickness. A heat pulse with an amplitude of 0.9 W and a duration of 0.12 s is applied to each discrete element. Through transient thermal analysis and structural mechanics analysis, the displacement response of each monitoring node of the entire frame crossbeam within 0-10 s after the heat input is recorded. A thermal sensitivity matrix Sx,y is constructed based on the response data. The monitoring nodes are uniformly distributed on the surface of the frame crossbeam, with a node spacing of 4 mm.

[0071] The iterative optimization parameters of the genetic algorithm are set as follows: initial population size of 80, search range of heat source location is within 50mm of the welding area of ​​the frame crossbeam, power search range is from 0.3 times the main welding power to the main welding power, and heating time sequence search step size is 0.1s; the number of iterations is set to 40 generations, crossover probability is 0.7, mutation probability is 0.08, and convergence of the objective function is used as the iteration termination condition. Finally, the optimal auxiliary heat source layout scheme is output.

[0072] During the welding process, the electromagnetic lock 522 of the flexible tooling platform 5 is released, causing the ball head support rod 521 to shift with the thermal deformation of the frame crossbeam, avoiding the influence of rigid constraints on the release of thermal stress, and further ensuring the welding deformation control effect.

[0073] The above are all preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Therefore, all equivalent changes made in accordance with the structure, shape and principle of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A multi-station collaborative welding method for automotive parts, characterized in that, Includes the following steps: Step 1: Construct a digital twin model of the thermal sensitivity of the welded workpiece to obtain the correlation between the workpiece's thermal input and deformation response; Step 2: Based on the thermal sensitivity digital twin model, the auxiliary heat source layout scheme is solved by optimization algorithm to form a coupled layout of the main welding heat source and the auxiliary stress control heat source, so that the auxiliary heat source generates a counteracting stress field with the opposite phase to the welding thermal deformation stress. Step 3: Assign at least two six-axis industrial robots (1) to perform the main welding task and the auxiliary stress control task respectively, and dynamically offset the welding deformation through the stress wave interference generated by the two heat sources; Step 4: Real-time acquisition of stress wave phase information through online monitoring device, and dynamic adjustment of auxiliary heat source output phase using phase-locked loop technology to ensure that the stress waves of the two heat sources always maintain anti-phase interference; Step 5: Calculate the workpiece thermal equilibrium point position in real time, and dynamically switch the roles of the main and auxiliary six-axis industrial robots (1) according to the thermal equilibrium point migration to ensure that the welding task is performed by the heat source that is most effective in controlling deformation.

2. The multi-station collaborative welding method for automotive parts according to claim 1, characterized in that, In step 1, the digital twin model of thermal sensitivity is constructed through finite element modal analysis, with the core parameter being the thermal sensitivity matrix. , defined as any discrete point on the workpiece surface Under a unit heat input, the displacement response matrix of each node of the workpiece is expressed as: ； in, This represents the total number of discrete units on the workpiece surface. This represents the total number of monitoring nodes for the entire workpiece. For the first The coordinates are When a unit heat pulse is applied to the discrete unit, the first The displacement response generated by each monitoring node.

3. The multi-station collaborative welding method for automotive parts according to claim 2, characterized in that, The method of finite element modal analysis is as follows: using finite element simulation software, the workpiece is discretized into several tetrahedral elements, and the element size is set to 0.5mm-2mm according to the workpiece wall thickness. A heat pulse with an amplitude of 0.8W-1W and a duration of 0.1-0.15s is applied to each discrete unit. Through transient thermal analysis and structural mechanical analysis, the displacement response of each monitoring node of the entire workpiece within 0-10s after the heat input is recorded. A thermal sensitivity matrix is ​​constructed based on the response data. The monitoring nodes are evenly distributed on the surface of the workpiece, with a node spacing of no more than 5 mm.

4. The multi-station collaborative welding method for automotive parts according to claim 3, characterized in that, The optimization algorithm in step 2 is a genetic algorithm, with the objective function being minimizing post-weld residual deformation. It iterative optimization is performed using the location, power, and heating sequence of the auxiliary heat source as variables. The objective function expression is: ; in, This represents the total residual deformation after welding. The predicted deformation amount generated by the main welding heat source at the j-th monitoring node. To compensate for the deformation generated by the heat source at the j-th monitoring node, a thermodynamic sensitivity matrix is ​​introduced as a weighting factor in the fitness function.

5. The multi-station collaborative welding method for automotive parts according to claim 4, characterized in that, The iterative optimization process of the genetic algorithm is as follows: the population size is initialized to 50-100, the search range of the heat source location is the welding area of ​​the workpiece and within 50mm of its perimeter, the power search range is 0.3 times the main welding power to the main welding power, the heating time sequence search step size is 0.1s, the number of iterations is set to 30-50 generations, the crossover probability is 0.6-0.8, the mutation probability is 0.05-0.1, and the convergence of the objective function is used as the iteration termination condition.

6. The multi-station collaborative welding method for automotive parts according to claim 5, characterized in that, The stress wave interference in step 3 is specifically implemented as follows: the main welding heat source operates at a constant power. Welding, with a frequency of Thermal stress waves; auxiliary heat source with power TIG remelting scanning is performed in a specific area of ​​the workpiece, and the movement speed of the auxiliary heat source or the frequency of the arc pulse is controlled to adjust the frequency of the generated thermal stress wave. satisfy ,in It is a frequency fine-tuning amount of 0.1-1Hz, used to adapt to dynamic phase drift.

7. The multi-station collaborative welding method for automotive parts according to claim 6, characterized in that, The TIG remelting scanning parameters for the auxiliary heat source are: scanning speed 5mm / s-15mm / s, arc voltage 10V-15V, and argon protective gas flow rate 8L / min-12L / min; under arc heating mode, the arc pulse frequency and thermal stress wave frequency are... Consistent, with a pulse duty cycle of 30%-50%.

8. The multi-station collaborative welding method for automotive parts according to claim 7, characterized in that, The online monitoring device in step 4 is a laser Doppler vibrometer array, used to calculate the instantaneous phase of the stress wave on the workpiece surface in real time. Phase-locked loop (PLL) technology uses a closed-loop system consisting of phase detection, loop filtering, and voltage-controlled oscillation to adjust the inverter frequency of the auxiliary heat source power supply, forcing the auxiliary heat source to output a different phase. satisfy To achieve anti-phase synchronization; The method for phase-locked loops is as follows: calculate the measured phase. The difference between the phase and the theoretical output Using a PI controller Filtering generates the control voltage U(t), with a PI controller proportional coefficient of 0.5-2.0 and an integral coefficient of 0.1-0.5; the inverter frequency of the auxiliary heat source power supply is adjusted via a numerically controlled oscillator until... Phase locking is achieved.

9. A multi-station collaborative welding method for automotive parts according to claim 8, characterized in that, The thermal equilibrium point in step 5 is the instantaneous neutral surface B(t) where the internal stress of the workpiece is zero, i.e. The contour lines are obtained; the local elastic modulus of the workpiece is updated based on the current temperature field, the thermal sensitivity matrix is ​​recalculated, and efficiency is adjusted. , ;in The displacement of the thermal equilibrium point caused by a unit of heat. Heat is input to the heat source; when the control efficiency of the original welding six-axis industrial robot (1) is lower than the threshold of 0.3mm / kJ-0.5mm / kJ, and the efficiency of another six-axis industrial robot (1) is higher than the threshold, the role swap is triggered.

10. A multi-station collaborative automotive parts welding device, characterized in that, To implement the multi-station collaborative automotive parts welding method of claim 8, the apparatus includes at least two six-axis industrial robots (1), welding torches (2) matching the number of six-axis industrial robots (1), a welding controller (3), a laser Doppler vibration meter array (4), a flexible tooling platform (5), and a three-dimensional structured light scanner (6). The welding torches (2) are detachably mounted on the movable end of the six-axis industrial robots (1). The welding controller (3) includes a thermo-coupling simulation module (31), a genetic algorithm optimization module (32), and a digital phase-locked loop controller (33). The thermo-coupling simulation module (31) is used to construct a thermo-sensitive digital twin model and calculate and predict the deformation field. The genetic algorithm optimization module (32) is used to solve the auxiliary heat source layout scheme. The digital phase-locked loop controller is used to realize the real-time synchronization of the auxiliary heat source phase. The laser Doppler... The vibration meter array (4) is arranged in a mesh above the welding station to collect the stress wave propagation speed, phase and amplitude information of the workpiece surface in real time without contact. The flexible tooling platform (5) is used to support the workpiece. The flexible tooling platform (5) includes a support base plate (51) and several sets of universal adaptive support units. The universal adaptive support unit includes a ball head support rod (521) and an electromagnetic lock (522). During the welding process, the electromagnetic lock (522) is released, so that the ball head support rod (521) is displaced with the thermal deformation of the workpiece. The three-dimensional structure light scanner (6) is installed on the top of the flexible tooling platform (5) to acquire the three-dimensional point cloud data of the workpiece after welding, and to detect residual deformation and feedback optimization. The welding controller (3) is connected to the six-axis industrial robot (1), welding torch (2), laser Doppler vibration meter array (4) and three-dimensional structure light scanner (6) respectively.

Images