Intelligent control method and system for industrial robot welding

Abstract

This invention belongs to the field of automatic arc metal cutting and welding equipment manufacturing technology, specifically involving an intelligent control method and system for industrial robot welding. The method includes: acquiring material property parameters and physical boundary dimensions of thin-walled sheet metal parts; synchronously collecting real-time arc voltage, real-time welding current, and actual feed speed during the robot's welding operation along a planned trajectory; calculating the out-of-plane thermal buckling bulge height of the thin-walled sheet metal part in the normal direction based on the collected parameters; and superimposing the bulge height as a feedforward compensation value into the vertical axis kinematic model to control the robot's welding torch to instantaneously lift at the same height in the normal direction. This invention utilizes electrical parameters with high-frequency response characteristics to replace lagging visual observation, achieving preventative displacement compensation for local deformation directly below the arc, effectively solving the burn-through problem in the welding process of thin-walled sheet metal parts, maintaining arc stability, and significantly improving welding yield.

Description

Technical Field

[0001] This invention relates to the field of manufacturing technology for automated arc metal cutting and welding equipment. More specifically, this invention relates to intelligent control methods and systems for industrial robot welding. Background Technology

[0002] In the field of electrical component manufacturing, robots are often used for welding thin-walled sheet metal parts of home appliances such as refrigerators and freezers. Because the sheet metal parts such as the outer shell of home appliances are thin, the concentrated heat input during the welding process can easily cause deformation of the sheet metal. In order to ensure the quality and forming effect of the weld, the welding process needs to be precisely controlled.

[0003] Chinese patent document CN103418942B discloses an intelligent control method for a welding robot. The method involves setting up multiple workstations within a workstation. When a weld point becomes difficult to weld at one workstation and requires turntable rotation, the welding robot first moves to another workstation to perform the welding operation. After the turntable at the original workstation rotates to its correct position, the robot returns to the original workstation to continue working, recording the interruption location for subsequent continuation. However, this patent document primarily addresses the logical sequence of multi-workstation switching to reduce the robot's waiting time during turntable rotation. Its focus is on the macro-level scheduling of production rhythm, without addressing how to handle the physical position fluctuations caused by the heating of thin plates during specific welding operations.

[0004] Chinese patent document CN114952098B discloses an intelligent weld seam positioning system and welding robot based on machine vision guidance. The system uses a camera in a preprocessing unit to scan and photograph the workpiece surface, a signal processor to identify the weld seam and calculate the coordinates of its start and end points, generating teaching information to guide the welding torch movement, and ultrasonic flaw detectors to inspect quality after welding. However, this patent relies heavily on image acquisition and pre-calculation before welding for positioning and path planning. Although a miniature industrial camera is used to further confirm the position, in actual welding, thin plates heat up extremely quickly and have a small heat capacity, causing them to momentarily bulge upwards under the heat of the electric arc. Due to the extremely high brightness of the welding arc, the camera cannot directly see the real-time bulging of the plate directly below the arc. This vision-based feedback method has a significant lag and cannot adjust the welding torch's position in real time according to the real-time bulging height of the plate.

[0005] In existing technologies, although some solutions attempt to improve efficiency through multi-station scheduling or utilize visual recognition technology for weld seam positioning, these solutions mostly focus only on the static position of the weld seam in a plane or macroscopic path planning. Due to the physical characteristics of thin-walled sheet metal parts for household appliances, when welding heat is continuously injected, the pressure generated by the thermal expansion of the sheet metal, after being restricted by the surrounding tooling fixtures, can only be released in a direction perpendicular to the sheet surface, resulting in sudden bulging or wavy deformation of the sheet metal. Most existing control methods move according to a pre-set height or based on visual characteristics of a previous distance, failing to capture the instantaneous deformation occurring directly below the arc. This causes the actual distance between the welding torch and the workpiece surface to fluctuate, affecting not only the stability of the arc but also, in cases of severe bulging, even causing the welding torch to directly contact and burn through the thin sheet, thus affecting the stability of welding thin-walled sheet metal parts. Summary of the Invention

[0006] To address the technical problem that transient bulging deformation caused by heat during the welding process of thin-walled sheet metal parts leads to uncontrolled distance between the welding torch and the workpiece surface under existing control methods, which easily causes burn-through, the present invention provides solutions in the following aspects.

[0007] In a first aspect, the present invention provides an intelligent control method for industrial robot welding, comprising: S1: acquiring the material property parameters and physical boundary dimensions of the thin-walled sheet metal part at the current workstation; the material property parameters include the coefficient of linear expansion, specific heat capacity, and density; the physical boundary dimensions include the thickness of the thin-walled sheet metal part and the lateral constraint span; S2: during the welding operation performed by the robot along the planned trajectory, synchronously acquiring the real-time arc voltage, real-time welding current, and actual feed speed at a preset sampling frequency; S3: calculating the out-of-plane thermal buckling bulge height in the normal direction of the thin-walled sheet metal part based on the real-time arc voltage, real-time welding current, actual feed speed, material property parameters, and physical boundary dimensions; S4: superimposing the out-of-plane thermal buckling bulge height as a feedforward compensation value into the vertical axis kinematic model of the industrial robot, and controlling the robot welding torch to instantaneously lift to the same height in the normal direction.

[0008] This invention, after acquiring the inherent material properties and physical boundary dimensions of thin-walled sheet metal parts, synchronously collects real-time arc voltage, real-time welding current, and actual feed speed during the welding process. Using these electrical and kinematic parameters, the out-of-plane thermal buckling bulge height in the normal direction of the thin-walled sheet metal part is directly calculated. This height is then used as a feedforward compensation value and superimposed on the vertical axis kinematic model to control the instantaneous lifting of the welding torch. This method directly addresses the underlying thermodynamic energy conversion mechanism and the geometric constraints of structural deformation. It utilizes electrical parameters with high-frequency response characteristics to replace the lag in visual sensor observations. Before physical deformation causes burn-through of the thin-walled sheet metal part, preventative displacement compensation for localized deformation directly below the arc is achieved. The control mechanism cuts off the vicious transmission chain caused by the shortened arc spacing due to the bulge of the thin-walled sheet metal part, leading to a surge in current. This ensures that the distance between the contact nozzle and the surface of the real-time bulging thin-walled sheet metal part remains constant, effectively solving the problem of burn-through caused by transient thermal buckling during continuous welding of thin-walled sheet metal parts and maintaining the physical stability of arc combustion.

[0009] Preferably, the step of calculating the out-of-plane thermally induced buckling height of the thin-walled sheet metal part in the normal direction based on the real-time arc voltage, real-time welding current, actual feed speed, material property parameters, and physical boundary dimensions includes: calculating the heat input per unit length based on the real-time arc voltage, real-time welding current, actual feed speed, and a preset effective arc thermal efficiency coefficient; and calculating the equivalent thermally induced temperature rise term based on the heat input per unit length, specific heat capacity, density, thickness, and a preset effective heat-affected zone width.

[0010] For calculating the out-of-plane thermally induced buckling height of thin-walled sheet metal parts, this invention first calculates the heat input per unit length based on electrical and kinematic parameters, and then calculates the equivalent thermally induced temperature rise term by combining the specific heat capacity and density of the material. This operation transforms the fluctuation of electrical parameters and the mechanical motion speed into specific local heat injection indicators, and further objectively maps the heat accumulation into the local temperature jump amplitude of the thin-walled sheet metal part. This energy conservation derivation method enables the control unit to grasp the local thermodynamic state changes under the impact of high-energy heat sources in real time and accurately, providing a temperature rise data benchmark for subsequent calculation of physical expansion, and improving the physical accuracy of deformation prediction.

[0011] Preferably, the step of calculating the heat input per unit length based on the real-time arc voltage, real-time welding current, actual feed speed, and preset arc effective thermal efficiency coefficient includes: multiplying the arc effective thermal efficiency coefficient, real-time arc voltage, and real-time welding current to obtain the net input power; and dividing the net input power by the actual feed speed to obtain the heat input per unit length.

[0012] Preferably, the step of calculating the equivalent thermal temperature rise term based on the heat input per unit length, specific heat capacity, density, thickness, and the preset effective heat-affected zone width includes: dividing the heat input per unit length by the product of the specific heat capacity, thickness, density, and the preset effective heat-affected zone width to obtain the equivalent thermal temperature rise term.

[0013] Preferably, after calculating the equivalent thermally induced temperature rise term, the method further includes: multiplying the lateral constraint span, the linear expansion coefficient, and the equivalent thermally induced temperature rise term to obtain the constrained expansion increment.

[0014] To derive the constrained expansion increment, this invention calculates the product of the lateral constraint span, the coefficient of linear expansion, and the equivalent thermal temperature rise term. This processing method utilizes the basic principle of linear thermal expansion to directly convert the local temperature field change into the material's free elongation trend within a specific constraint range. By introducing the inherent sensitivity index of the coefficient of linear expansion, the algorithm can accurately respond to the differences in the physical expansion and contraction characteristics of different thin-walled sheet metal materials when heated. This allows the controller to quickly identify the magnitude of the horizontal internal stress caused by the internal lattice expansion, thereby providing reliable intermediate geometric parameters for the subsequent calculation of the three-dimensional out-of-plane bulge height.

[0015] Preferably, after obtaining the restricted expansion increment, the method further includes: multiplying the lateral constraint span by the restricted expansion increment to obtain the distortion displacement modulus; and performing a square root algebra operation on the distortion displacement modulus to obtain the out-of-plane thermal buckling bulge height.

[0016] After obtaining the constrained expansion increment, this invention multiplies it by the transverse constraint span to obtain the distortion displacement modulus, and then uses square root operation to obtain the final out-of-plane thermal buckling bulge height. This operation utilizes the spatial geometric compression relationship of structural buckling under compression, removes high-order invalid calculation terms, and effectively amplifies the tiny invisible expansion of the trapped metal in the horizontal plane into a visible physical bulge in the vertical direction according to the physical conduction path of least resistance. This significantly reduces the computational burden of the underlying control chip, achieves millisecond-level calculation of geometric deformation, and ensures the real-time issuance of feedforward control commands.

[0017] Preferably, the process of obtaining the material property parameters and physical boundary dimensions of the thin-walled sheet metal part at the current workstation includes: reading the linear expansion coefficient, specific heat capacity, and density pre-calibrated and stored in the memory through the control unit; obtaining the thickness and lateral constraint span of the thin-walled sheet metal part at the current workstation through a non-contact measuring mechanism; and writing the linear expansion coefficient, specific heat capacity, density, thickness, and lateral constraint span as discrete constants into the operation register of the underlying controller.

[0018] Preferably, the real-time synchronous acquisition of the current arc voltage, real-time welding current, and actual feed speed at a preset sampling frequency includes: intercepting the real-time arc voltage through a potential sensing branch deployed at the output end of the welding power source; sensing the real-time welding current using the Hall effect principle; and calculating the actual feed speed of the tool center point in real time based on the displacement vectors of each joint fed back by the encoder through the robot spindle controller.

[0019] Preferably, the step of superimposing the out-of-plane thermal buckling bulge height as a feedforward compensation value into the vertical axis kinematic model of the industrial robot, and controlling the robot welding torch to instantaneously lift to the same height in the normal direction, includes: converting the calculated out-of-plane thermal buckling bulge height into the normal displacement increment of the robot end effector; superimposing the calculated out-of-plane thermal buckling bulge height into the current path planning coordinates in real time through the motion control kernel of the industrial robot; when the out-of-plane thermal buckling bulge height is a positive value, controlling the coordinated movement of the servo motors of each joint through the underlying driver to drive the welding torch to instantaneously lift to the same physical height in the normal direction to avoid collision.

[0020] When performing feedforward compensation, this invention converts the calculated out-of-plane thermal buckling bulge height into a normal displacement increment and superimposes it in real time into the path planning coordinates to drive the servo motor to lift the welding torch. This control mode changes post-event position correction to pre-event dynamic avoidance. By directly intervening in the underlying driver, it forcibly endows the mechanical actuator with microsecond-level terrain undulation following ability. This advanced command logic opens up a safe distance between the conductive nozzles in advance before the thin-walled sheet metal part actually suffers catastrophic burn-through, effectively mitigating the risk of hard contact caused by thermal buckling extrusion and ensuring the smooth operation of the entire welding metallurgical cycle.

[0021] In a second aspect, the present invention provides an intelligent control system for industrial robot welding, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned intelligent control method for industrial robot welding is implemented.

[0022] By adopting the above technical solution, the intelligent control method for industrial robot welding is generated into a computer program and stored in a memory for loading and execution by a processor. This allows for the creation of terminal devices based on the memory and processor, facilitating their use.

[0023] The beneficial effects of this invention are as follows: This invention, after acquiring the inherent material properties and physical boundary dimensions of thin-walled sheet metal parts, synchronously collects real-time arc voltage, real-time welding current, and actual feed speed during the welding process. Using these electrical and kinematic parameters, the out-of-plane thermal buckling bulge height in the normal direction of the thin-walled sheet metal part is directly calculated. This height is then used as a feedforward compensation value and superimposed on the vertical axis kinematic model to control the instantaneous lifting of the welding torch. This method directly addresses the underlying thermodynamic energy conversion mechanism and the geometric constraints of structural deformation. It utilizes electrical parameters with high-frequency response characteristics to replace the lag in visual sensor observations. Before physical deformation causes burn-through of the thin-walled sheet metal part, preventative displacement compensation for localized deformation directly below the arc is achieved. The control mechanism cuts off the vicious transmission chain caused by the shortened arc spacing due to the bulge of the thin-walled sheet metal part, leading to a surge in current. This ensures that the distance between the contact nozzle and the surface of the real-time bulging thin-walled sheet metal part remains constant, effectively solving the problem of burn-through caused by transient thermal buckling during continuous welding of thin-walled sheet metal parts and maintaining the physical stability of arc combustion. Attached Figure Description

[0024] Figure 1 A flowchart of the intelligent control method for industrial robot welding in this invention is shown; Figure 2 A schematic diagram of the internal structure of the out-of-plane thermally induced buckling height calculation module is shown. Detailed Implementation

[0025] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] To address the transient failure challenges faced by thin-walled sheet metal parts in automated welding processes, this embodiment establishes real-time avoidance logic by deeply analyzing the physical coupling mechanism between arc heat flow and the geometric deformation of thin-walled sheet metal parts. When encountering the severe thermal expansion induced by the extremely small heat capacity of thin-walled sheet metal parts, this invention abandons the traditional visual sensor's delayed observation of the square variable behind the molten pool and instead utilizes electrical monitoring parameters with high-frequency response characteristics. The core mechanism of this design is to convert the current feedback signal used for stabilizing the arc length control loop into geometric perception calculations. By actively avoiding dependence on complex finite element simulations of the global thermal strain field, this scheme concentrates redundant global computing power into millisecond-level response characteristics to local deformation directly below the welding torch, enabling the robot to perform preventative displacement compensation based on energy balance principles before physical deformation leads to burn-through.

[0027] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0028] This invention discloses an intelligent control method for industrial robot welding, referring to... Figure 1 This includes steps S1-S4: S1. Obtain the parameters and physical boundary dimensions of the thin-walled sheet metal part.

[0029] It should be noted that accurately reflecting the thermal buckling phenomenon requires locking the constitutive characteristics and spatial mechanical constraint range of the thin-walled sheet metal material: the degree of lattice expansion of the thin-walled sheet metal when heated is determined by its inherent thermophysical constants, while the lateral pressure applied by the tooling fixture is the direct driving force for the thin-walled sheet metal to bulge in the normal direction; this embodiment defines the physical boundary of strain energy release by introducing a lateral constraint span. This parameter configuration method aims to establish a stable geometric coordinate reference for subsequent analytical calculations, so as to ensure that the prediction model can adapt to thin-walled sheet metal parts of different specifications. When facing material changes caused by production line changes, this solution can maintain the robustness of the underlying control mechanism by adjusting the input parameters.

[0030] Specifically, the control unit reads the linear expansion coefficient that has been pre-calibrated and stored in the memory. Specific heat capacity and density The thickness of the thin-walled sheet metal part at the current workstation is obtained through a non-contact measuring mechanism deployed on-site. and lateral constraint span The coefficient of linear expansion read Specific heat capacity ,density ,thickness and lateral constraint span All of them are written as discrete constants into the arithmetic registers of the underlying controller.

[0031] Among them, the coefficient of linear expansion The sensitivity to changes in temperature field and geometric length; specific heat capacity With density The product determines the local thermal inertia of thin-walled sheet metal parts, and the lateral constraint span This limits the range of compressive resistance of the heated metal material in the horizontal plane; since the physical properties provided by the manufacturer are mostly static nominal values ​​at room temperature, they cannot characterize the nonlinear changes of the material under extreme arc thermal gradients; all parameters need to be dynamically optimized and calibrated in advance based on measured data from actual high-temperature welding environments in the laboratory, so as to transform the material parameters into system equivalent parameters that incorporate the heat dissipation characteristics on site, thereby effectively reducing the cumulative calculation noise caused by factory tolerances and high-temperature nonlinear drift in the subsequent multiplication chain.

[0032] S2. Real-time acquisition of welding electrical parameters and actual feed speed.

[0033] It should be noted that fluctuations in heat input intensity are the direct cause of the instantaneous abrupt change in bulge displacement. Since the distance between the electric arc and the thin-walled sheet metal part changes drastically with the bulge of the thin-walled sheet metal part, real-time monitoring of the instantaneous transitions of electrical parameters becomes a key means of capturing geometric distortion. This invention transforms the invisible energy injection process into a corresponding digital feature stream by high-frequency interception of voltage and current data. To combat electromagnetic pulse interference generated by the high-frequency inverter power supply, a digital smoothing mechanism is built into the sampling link. This data processing strategy actively filters out the glitches of extremely high-frequency waveforms, thereby establishing the signal stability that characterizes the main trend of heat input, thus providing a clean physical driving signal for subsequent geometric calculation.

[0034] Specifically, the sampling unit intercepts the real-time arc voltage through a potential sensing branch deployed at the output of the welding power source. Simultaneously utilize the Hall effect principle to sense real-time welding current. The robot spindle controller calculates the actual feed rate of the tool center point in real time based on the displacement vectors of each joint fed back by the encoder. .

[0035] Among them, the collected real-time arc voltage Real-time welding current and actual feed rate Updated at kilohertz frequency; real-time arc voltage With real-time welding current The product is directly mapped to the instantaneous output thermal power, while the actual feed rate... This represents the integral density of the power in space. This multi-dimensional data acquisition architecture ensures that the system can perceive the precise energy value received over every millimeter of physical length in the welding heat-affected zone in real time, providing underlying engineering support for solving the problem of deformation hysteresis prediction.

[0036] S3. Calculate the out-of-plane thermal buckling bulge height of the thin-walled sheet metal part.

[0037] It should be noted that the evolution from electrical energy injection to mechanical displacement follows strict thermodynamic equilibrium and geometric compatibility principles: when a thin-walled sheet metal part is subjected to a high-energy electric arc impact, the local temperature rise rapidly exceeds its material yield threshold. Since it is rigidly fixed by the rigid tooling within the transverse constraint span, the horizontal elongation trend is transformed into normal structural buckling. In this embodiment, a nonlinear mapping model based on the Euler buckling principle and the law of thermal expansion and contraction is selected. The motivation for this selection is that the model can express the nonlinear thermally induced bulging physical process in a concise analytical structure, thereby meeting the strict requirements of the underlying controller for computational efficiency. Specifically, the electrical parameter input is first converted into a local temperature rise, then the temperature rise is combined with the linear expansion coefficient to calculate the physical deformation, and finally the horizontal deformation is converted into a vertical bulge height through the transverse constraint span. In this parameter closed loop, the microscopic effect is transformed into a visible out-of-plane normal displacement, realizing a direct numerical mapping from data to deformation.

[0038] Specifically, based on the law of conservation of energy and the geometric compression relationship of isosceles triangles, the out-of-plane thermally induced buckling height is constructed. The derivation and calculation process is as follows: According to basic electrical principles, the transient thermal power of a moving electric arc input to a thin-walled sheet metal part is equal to the real-time arc voltage. With real-time welding current The product of, i.e. Combined with the effective thermal efficiency coefficient of the electric arc during energy transfer. The net input power entering the surface of the thin-walled sheet metal part is .

[0039] In actual feed rate Within the physical scale, the heat received per unit length of a thin-walled sheet metal part is the heat input per unit length. The following state equation is satisfied:

[0040] In the formula, Represents heat input per unit length, measured in joules per meter; This represents the effective thermal efficiency coefficient of the electric arc, and is a dimensionless constant. Represents the real-time arc voltage, measured in volts; This represents the real-time welding current, measured in amperes. This represents the actual feed rate, measured in meters per second.

[0041] According to the thermodynamic law of conservation of energy, the heat input per unit length Injection inevitably causes a temperature rise in localized thin-walled sheet metal parts, and the amount of heat absorbed is equal to the mass of metal per unit length that participates in heat absorption. Specific heat capacity With equivalent heat-induced temperature rise term The product of consecutive products, i.e. Within a unit physical length, the physical volume of a thin-walled sheet metal part that participates in heat absorption is equal to its thickness. , preset effective heat-affected zone width And the product with the scalar 1 per unit length, thus deriving the amount of heat absorbed per unit length. equal to density ,thickness With respect to the preset effective heat-affected zone width The product of, i.e. .

[0042] The physical essence of the preset effective heat-affected zone width is the spatial span through which the high-energy heat of the electric arc is conducted to the metal lattice and causes significant thermal expansion. Obtaining this parameter depends on the actual heat transfer boundary test under the benchmark calibration conditions. Specifically, before the formal operation, thin-walled sheet metal parts of the same batch are sampled for rated parameter test welding. During this benchmark test welding process, a high-precision infrared thermal imager is used to simultaneously scan the dynamic surface temperature field distribution on both sides of the weld. The test platform extracts the absolute distance of the lateral span above the dynamic yield critical temperature of the thin-walled sheet metal material on the temperature gradient curve. After multiple multi-point measurements to obtain the arithmetic mean, the control unit writes the arithmetic mean as the preset effective heat-affected zone width into the operation register of the underlying controller.

[0043] The amount of heat absorbed per unit length Substituting the algebraic relations into the thermodynamic energy conservation equation, we get This leads to the derivation of the equivalent heat-induced temperature rise term. The following state equation is satisfied:

[0044] In the formula, This represents the equivalent thermally induced temperature rise, expressed in Kelvin. Represents heat input per unit length, measured in joules per meter; Specific heat capacity is expressed in joules per kilogram (Kelvin). Represents density, expressed in kilograms per cubic meter; Represents thickness, in meters; This represents the preset effective heat-affected zone width, in meters.

[0045] Based on the fundamental principle of linear thermal expansion, the transverse constraint span... Limited expansion increment of thin-walled sheet metal parts under heat within the specified range With linear expansion coefficient and equivalent heat-induced temperature rise term They exhibit a linear proportional relationship and satisfy the following dimensional equation: ; heat input per unit length and equivalent heat-induced temperature rise After substituting the algebraic expression, the constrained inflation increment is derived. The following dimensional equations must be satisfied:

[0046] In the formula, Represents the constrained expansion increment, in meters; This represents the lateral constraint span, in meters. This represents the coefficient of linear expansion, expressed in Kelvin. This represents the effective thermal efficiency coefficient of the electric arc, and is a dimensionless constant. Represents the real-time arc voltage, measured in volts; This represents the real-time welding current, measured in amperes. This represents the actual feed rate, measured in meters per second. Specific heat capacity is expressed in joules per kilogram (Kelvin). Represents density, expressed in kilograms per cubic meter; This represents the preset effective heat-affected zone width, in meters. This represents thickness, measured in meters.

[0047] Because the thin-walled sheet metal parts are rigidly fixed on both sides by matching clamps, the lateral constraint span before and after heating is... When kept constant on a horizontal two-dimensional surface, the extra microscopic physical length due to thermal expansion cannot extend laterally within the two-dimensional plane. Consequently, the internal stress inevitably forces the thin-walled sheet metal part in the central region to undergo outward buckling deformation.

[0048] The physical profile of the raised section is simplified into an approximately isosceles triangular abrupt geometric structure, and then perpendicularly cut along the central axis to obtain a right-angled triangular model for calculation. In this right-angled triangular model, the lateral constraint span is... Half of the distance forms the bottom edge, and the length of the half-section of the solid after thermal expansion forms the hypotenuse of the right triangle. The height of the out-of-plane thermal buckling bulge forms the right-angled side of this right triangle. Because the thin-walled sheet metal part within the span absorbs heat and undergoes linear thermal expansion, the physical length of the hypotenuse is equal to the transverse constraint span. Half of the limited expansion increment The sum of half of them.

[0049] Based on the spatial geometric relationships of the Pythagorean theorem, the following algebraic equations for the legs and hypotenuses of a right triangle are constructed:

[0050] Expanding and simplifying the squared terms on the right-hand side of the above algebraic equation, we derive the following polynomial equation:

[0051] By simultaneously eliminating the constant terms of the same name from both sides of the above polynomial equation, the following equation is derived:

[0052] In objective engineering reality, due to the limited expansion increment The physical value is much smaller than the lateral constraint span. The initial length, the constrained expansion increment The higher-order squared terms are ignored in engineering mathematical calculations; at the same time, in order to simplify the feedforward control computational complexity of the underlying controller and to broadly cover spatial nonlinear bulge conditions, proportional approximation mapping logic is directly used for coefficient normalization, deriving the following nonlinear spatial mapping equation:

[0053] By simultaneously performing square root algebraic operations on both sides of the above nonlinear spatial mapping equation, the out-of-plane thermally induced buckling height is derived. Physical mapping relationship:

[0054] In the formula, Represents the height of out-of-plane thermally induced buckling bulge, in meters; This represents the lateral constraint span, in meters. Represents the limited expansion increment, in meters.

[0055] constrained expansion increment Substituting the complete algebraic expression into the above physical mapping relationship and performing algebraic substitution, the final out-of-plane thermally induced buckling height is constructed. The algebraic mapping relationship is as follows:

[0056] In the formula, Represents the height of out-of-plane thermally induced buckling bulge, in meters; This represents the lateral constraint span, in meters. This represents the coefficient of linear expansion, expressed in Kelvin. This represents the effective thermal efficiency coefficient of the electric arc, and is a dimensionless constant. Represents the real-time arc voltage, measured in volts; This represents the real-time welding current, measured in amperes. This represents the actual feed rate, measured in meters per second. Specific heat capacity is expressed in joules per kilogram (Kelvin). Represents density, expressed in kilograms per cubic meter; This represents the preset effective heat-affected zone width, in meters. This represents thickness, measured in meters.

[0057] In analytical computing architectures It constitutes transient thermal power, which characterizes the magnitude of the instantaneous output physical thermal power; It constitutes the net input power, which characterizes the effective heat flux density entering the metal surface; The heat input per unit length is characterized by the total amount of heat energy retained by the electric arc along its trajectory within a unit physical scale. It constitutes an equivalent thermally induced temperature rise term, which characterizes the macroscopic temperature jump amplitude of the local welded area after being impacted by a high-energy pulsed heat source. The constrained expansion increment characterizes the tendency of thin-walled sheet metal parts to freely elongate under lateral physical constraints. The distortion displacement modulus is a squared characteristic value representing the change in out-of-plane geometry of a thin-walled sheet metal part at the instant of thermal buckling.

[0058] The specific operating mechanism of the calculation formula is as follows: when the real-time arc voltage... Or real-time welding current When subjected to disturbances and increased dramatically, the effective energy accumulation entering the thin-walled sheet metal increases, leading to a rise in the equivalent thermally induced temperature rise and an enhanced increase in confined expansion; due to the lateral constraint span Providing a rigid lateral reaction force, this linear expansion tendency is forced to propagate in the vertical direction of least resistance, resulting in an out-of-plane thermally induced buckling bulge height. Consequently, the actual feed rate increases. Under extreme conditions approaching stagnation, a sharp decrease in the denominator of the internal calculation leads to a sharp increase in the predicted bulge displacement. This dynamic feature accurately fits the real physical manifestation of the precursor to burn-through caused by local heat accumulation, ensuring that all electrical signal disturbances in the underlying data stream are accurately mapped to spatial geometric abrupt changes.

[0059] For example, Figure 2 The diagram shows the internal structure of the out-of-plane thermally induced buckling bulge height calculation module, illustrating the internal data calculation chain that progressively deduces from electrical and kinematic parameters to net input power, heat input per unit length, equivalent thermally induced temperature rise term, constrained expansion increment, and finally to the output bulge height.

[0060] S4. Perform feedforward compensation to control the instantaneous lifting of the welding torch in the normal direction.

[0061] It should be noted that traditional trajectory tracking mechanisms typically perform mechanical compensation only after optical deviation occurs. This physical lag is unacceptable in scenarios where welding of thin-walled sheet metal parts results in extremely rapid thermal penetration failure. In this embodiment, the calculated displacement prediction value is directly injected into the underlying drive level of the motion controller, enabling proactive command of the mechanical actuator. This feedforward compensation logic cuts off the vicious physical chain that leads to arc length shortening due to deformation, current surge due to arc length shortening, and current surge exacerbating deformation, thereby forcibly stabilizing the welding process near the set process baseline.

[0062] Specifically, the computational unit will calculate the out-of-plane thermally induced buckling height. The normal displacement increment is converted into that of the robot's end effector; the industrial robot's motion control kernel will then calculate the out-of-plane thermally induced buckling height. The out-of-plane thermal buckling height is superimposed onto the current path planning coordinates in real time. When the value is positive, the bottom-level driver immediately controls the coordinated movement of the servo motors of each joint, causing the welding torch to instantly rise and avoid obstacles at the same physical height in the normal direction; the dynamic displacement adjustment action ensures that the distance between the conductive nozzle and the surface of the real-time raised thin-walled sheet metal part remains constant; this dynamic avoidance action is continuously and frequently executed during the continuous operation of the entire weld seam, so that the robot's movement trajectory is synchronized with the thermal morphological changes of the thin-walled sheet metal part. Through this physical-level dynamic adaptive adjustment, this embodiment maintains the physical stability of arc combustion under complex thermodynamic interference conditions, and cuts off the conditions that cause the generation of burn-through electrical peaks from the bottom layer.

[0063] This invention also discloses an intelligent control system for industrial robot welding, including a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, the intelligent control method for industrial robot welding according to this invention is implemented.

[0064] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.

Claims

1. An intelligent control method for industrial robot welding, characterized in that, include: S1: Obtain the material property parameters and physical boundary dimensions of the thin-walled sheet metal part at the current workstation; the material property parameters include the coefficient of linear expansion, specific heat capacity, and density; Physical boundary dimensions include the thickness of thin-walled sheet metal parts and the lateral constraint span; S2: During the welding operation performed by the robot along the planned trajectory, the real-time arc voltage, real-time welding current and actual feed speed are collected synchronously at a preset sampling frequency. S3: Calculate the out-of-plane thermal buckling bulge height of the thin-walled sheet metal part in the normal direction based on the real-time arc voltage, real-time welding current, actual feed speed, material property parameters, and physical boundary dimensions. S4: The out-of-plane thermal buckling height is superimposed as a feedforward compensation value into the vertical axis kinematic model of the industrial robot to control the robot's welding torch to instantaneously lift to the same height in the normal direction.

2. The intelligent control method for industrial robot welding according to claim 1, characterized in that, The calculation of the out-of-plane thermal buckling bulge height of the thin-walled sheet metal part in the normal direction based on real-time arc voltage, real-time welding current, actual feed speed, material property parameters, and physical boundary dimensions includes: The heat input per unit length is calculated based on the real-time arc voltage, real-time welding current, actual feed speed, and preset effective arc thermal efficiency coefficient. The equivalent thermal temperature rise term is calculated based on the heat input per unit length, specific heat capacity, density, thickness, and the preset effective heat-affected zone width.

3. The intelligent control method for industrial robot welding according to claim 2, characterized in that, The calculation of heat input per unit length based on real-time arc voltage, real-time welding current, actual feed speed, and preset effective arc thermal efficiency coefficient includes: The net input power is obtained by multiplying the effective thermal efficiency coefficient of the electric arc, the real-time arc voltage, and the real-time welding current. Divide the net input power by the actual feed rate to obtain the heat input per unit length.

4. The intelligent control method for industrial robot welding according to claim 2, characterized in that, The calculation of the equivalent thermal temperature rise term based on heat input per unit length, specific heat capacity, density, thickness, and the preset effective heat-affected zone width includes: Divide the heat input per unit length by the product of specific heat capacity, thickness, density and the preset effective heat-affected zone width to obtain the equivalent thermal temperature rise term.

5. The intelligent control method for industrial robot welding according to claim 4, characterized in that, After calculating the equivalent thermally induced temperature rise term, the method further includes: Multiplying the lateral constraint span, the linear expansion coefficient, and the equivalent thermal temperature rise term yields the constrained expansion increment.

6. The intelligent control method for industrial robot welding according to claim 5, characterized in that, After obtaining the constrained expansion increment, the method further includes: Multiply the lateral constraint span by the restricted expansion increment to obtain the distortion displacement modulus; The out-of-plane thermal buckling height is obtained by performing a square root algebraic operation on the distortion displacement modulus.

7. The intelligent control method for industrial robot welding according to claim 1, characterized in that, The process of obtaining the material property parameters and physical boundary dimensions of the thin-walled sheet metal part at the current workstation includes: The linear expansion coefficient, specific heat capacity, and density are read from the pre-calibrated and stored memory by the control unit. The thickness and lateral constraint span of the thin-walled sheet metal part at the current workstation are obtained through a non-contact measuring mechanism. The coefficient of linear expansion, specific heat capacity, density, thickness, and lateral constraint span are written as discrete constants into the operation register of the underlying controller.

8. The intelligent control method for industrial robot welding according to claim 1, characterized in that, The method of synchronously acquiring real-time arc voltage, real-time welding current, and actual feed speed at a preset sampling frequency includes: The real-time arc voltage is captured by a potential sensing branch deployed at the output end of the welding power source. Utilizing the Hall effect principle to sense real-time welding current; The robot spindle controller calculates the actual feed rate of the tool center point in real time based on the displacement vectors of each joint fed back by the encoder.

9. The intelligent control method for industrial robot welding according to claim 1, characterized in that, The step of adding the out-of-plane thermally induced buckling height as a feedforward compensation value to the vertical axis kinematic model of the industrial robot, and controlling the robot's welding torch to instantaneously lift by the same height in the normal direction, includes: The calculated out-of-plane thermal buckling height is converted into the normal displacement increment of the robot end effector; the calculated out-of-plane thermal buckling height is then superimposed onto the current path planning coordinates in real time through the industrial robot's motion control kernel; When the out-of-plane thermal buckling height is a positive value, the underlying driver controls the coordinated movement of the servo motors of each joint, causing the welding torch to be instantly raised to the same physical height in the normal direction to avoid collision.

10. An intelligent control system for industrial robot welding, characterized in that, include: A processor and a memory, the memory storing computer program instructions that, when executed by the processor, implement the intelligent control method for industrial robot welding according to any one of claims 1-9.

Images