Intelligent positioning and processing method and system for seat framework based on laser welding

Abstract

The application discloses a seat framework intelligent positioning processing method and system based on laser welding, relates to the related field of laser processing, and comprises the following steps: mining a main stress transmission path of a seat framework based on a scene working condition, and constructing a digital load atlas; determining a limiting strategy by two-stage decision of clamping limiting and clamping preloading application according to a positioning intelligent agent and taking the digital load atlas as a reference; performing assembly clamping limiting of parts of the seat framework, scanning and converting to determine a weld three-dimensional coordinate, combining a welding intelligent agent to determine welding driving parameters, and performing automatic welding driving on a laser welding assembly. The application solves the technical problem that existing seat framework laser welding exists independent clamping positioning and welding process, which leads to misplacement of a clamping point and a force flow key point, and further affects welding precision and structural reliability, realizes collaborative optimization of clamping limiting and welding driving, and further improves the technical effect of welding precision and structural reliability.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of laser processing, in particular to a seat framework intelligent positioning processing method and system based on laser welding. BACKGROUND

[0002] As the core bearing structure of the automobile / aviation seat, the welding quality of the seat framework directly affects the strength, durability and passenger safety of the seat, especially under complex working conditions, stress concentration at the weld joint is easy to cause structural fatigue failure, therefore, it is crucial to realize high-precision and high-reliability welding. At present, the mainstream method relies on experience to design fixture positioning scheme, and controls the robot for standardized welding through offline programming, and the welding parameters are usually selected based on the fixed process library. However, due to the complex structure of the seat framework and the variable working load, the traditional method is difficult to accurately match the actual stress distribution, resulting in the dislocation of the clamping positioning and the force flow transmission path, and after welding, problems such as residual stress concentration and deformation out-of-tolerance are easy to occur, and the static welding parameters cannot adapt to the fluctuation of the assembly gap, and the welding quality consistency is difficult to guarantee.

[0003] In the related art, the laser welding of the seat framework has the technical problem that the fixture positioning and the welding process are independent of each other, resulting in the dislocation of the clamping point and the key point of the force flow, and further affecting the welding precision and the structural reliability. SUMMARY

[0004] The present application provides a seat framework intelligent positioning processing method and system based on laser welding, which adopts the technical means of excavating the main stress transmission path of the seat framework to construct a digital load map marked with pressure-bearing and tension-bearing welds, determining the limiting strategy according to the two-stage decision of the positioning agent based on the digital load map, clamping and limiting according to the limiting strategy, scanning and converting to determine the three-dimensional coordinates of the welds, combining the welding agent to determine the parameters, and driving the laser welding assembly for automatic welding, etc. The technical problem that the fixture positioning and the welding process are independent of each other in the existing laser welding of the seat framework, resulting in the dislocation of the clamping point and the key point of the force flow, and further affecting the welding precision and the structural reliability is solved, the collaborative optimization of clamping and limiting and welding driving is realized, and the technical effects of improving the welding precision and the structural reliability are achieved.

[0005] The present application provides a seat framework intelligent positioning processing method based on laser welding, comprising: excavating the main stress transmission path of the seat framework based on the scene working condition, and constructing a digital load map, wherein the digital load map is marked with pressure-bearing welds and tension-bearing welds; according to the positioning agent, taking the digital load map as the reference, making two-stage decisions of clamping and limiting and clamping preloading, and determining the limiting strategy; assembling and clamping the parts of the seat framework according to the limiting strategy, scanning and converting to determine the three-dimensional coordinates of the welds, combining the welding agent to determine the welding driving parameters, and executing automatic welding driving on the laser welding assembly.

[0006] In a possible implementation, a digital load map is constructed, and the following processing is performed: constructing a lightweight assembly architecture of the seat framework according to a pre-processed part assembly and welding assembly requirements; performing stress cycle marking and weld type marking in the lightweight assembly architecture according to a main stress transmission path, to constitute the digital load map.

[0007] In a possible implementation, a positioning agent is constructed, and the following processing is performed: constructing a first positioning component based on clamping limiting and a second positioning component based on clamping preloading application, cascading the first positioning component and the second positioning component to generate a positioning decision architecture; embedding the digital load map in the positioning decision architecture and supervising training to convergence to generate the positioning agent.

[0008] In a possible implementation, clamping limiting decision is made, and the following processing is performed: positioning a key path and a position node of force flow transmission in a seat framework assembly according to the digital load map according to the first positioning component; determining primary and secondary positioning references according to the key path and the position node; making a rigid limiting application decision for the primary positioning reference to determine first clamping limiting, and making a flexible limiting application decision based on controllable degrees of freedom for the secondary positioning reference to determine second clamping limiting, wherein each clamping action point of the second clamping limiting is integrated with a precision actuator; and determining clamping limiting results according to the first clamping limiting and the second clamping limiting.

[0009] In a possible implementation, clamping preloading application decision is made, and the following processing is performed: determining a pre-displacement field, wherein the pre-displacement field is a pre-strain applied before welding; performing pre-strain parameter analysis and iterative optimization based on the pre-displacement field with the second clamping limiting as a clamping action point according to the second positioning component to determine flexible limiting parameters, wherein the flexible limiting parameters at least include force direction and displacement stroke of each action point; and generating clamping preloading application results according to the flexible limiting parameters.

[0010] In a possible implementation, the pre-displacement field is determined, and the following processing is performed: constructing a simulation field based on the clamping limiting results to simulate and determine an ideal residual stress distribution after welding; inversely calculating a first pre-strain to be applied before welding according to the residual stress distribution; scanning the digital load map to determine a second pre-strain based on working load; and coupling the first pre-strain and the second pre-strain to generate the pre-displacement field.

[0011] In a possible implementation, after the laser welding assembly is driven to perform automatic welding, the following processing is performed: an acoustic emitter is additionally installed on the end effector, wherein the acoustic emitter has a first time delay based on welding operation; the acoustic emitter performs welding energy efficiency detection along with the welding process of the laser welding assembly, and analyzes welding defects through acoustic emission signals, wherein the welding defects at least include pores and incomplete fusion defects; positioning and rework welding driving based on the welding defects are performed.

[0012] In a possible implementation, after the laser welding assembly is driven to perform automatic welding, the following processing is performed: an acoustic emitter is additionally installed on the end effector, wherein the acoustic emitter has a first time delay based on welding operation; the acoustic emitter performs welding energy efficiency detection along with the welding process of the laser welding assembly, and analyzes welding defects through acoustic emission signals, wherein the welding defects at least include pores and incomplete fusion defects; positioning and rework welding driving based on the welding defects are performed.

[0013] In a possible implementation, the following processing is performed: the laser welding assembly is a multi-axis assembly, when there is a welding defect, a first collaborative manipulator is driven; a rework positioning result is determined through positioning of the welding defect and three-dimensional coordinate conversion based on the first collaborative manipulator; the first collaborative manipulator is driven according to the rework positioning result to perform rework welding of the welding defect.

[0014] The application also provides a seat framework intelligent positioning processing system based on laser welding, comprising: a main stress transmission path excavation module, configured to excavate a main stress transmission path of a seat framework based on a scene working condition, and construct a digital load atlas, wherein the digital load atlas is marked with a pressure-bearing weld and a tension-bearing weld; a limiting strategy determination module, configured to determine a limiting strategy according to a positioning intelligent agent and based on the digital load atlas, and perform two-stage decision of clamping limiting and clamping preload application; and a welding driving module, configured to perform assembly clamping limiting of parts of the seat framework according to the limiting strategy, scan and convert to determine a weld three-dimensional coordinate, combine a welding intelligent agent to determine welding driving parameters, and drive a laser welding assembly to perform automatic welding.

[0015] The intelligent positioning processing method and system for a seat framework based on laser welding are proposed in the application. First, the main stress transmission path of the seat framework is excavated based on a scene working condition, and a digital load atlas is constructed, wherein the digital load atlas is marked with pressure-bearing welds and tension-bearing welds. Then, according to a positioning intelligent agent, two-stage decisions of clamping limiting and clamping preloading are made based on the digital load atlas, and a limiting strategy is determined. Finally, the seat framework is assembled and clamped according to the limiting strategy, the three-dimensional coordinates of the welds are scanned and converted, the welding driving parameters are determined in combination with a welding intelligent agent, and the automatic welding driving of the laser welding assembly is performed. The collaborative optimization of clamping limiting and welding driving is realized, and the technical effects of improving the welding precision and the structural reliability are achieved. BRIEF DESCRIPTION OF DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the application, the drawings of the embodiments of the application will be briefly introduced as follows. In the present application, a flowchart is used to illustrate the operations performed by the system according to the embodiments of the application. It should be understood that the foregoing or the following operations are not necessarily performed in sequence. On the contrary, various steps can be processed in reverse order or simultaneously according to needs. Meanwhile, other operations can be added to these processes, or a step or several steps can be removed from these processes.

[0017] Figure 1 A flowchart of the intelligent positioning processing method for a seat framework based on laser welding provided by the embodiments of the application is shown.

[0018] Figure 2 A structure diagram of the intelligent positioning processing system for a seat framework based on laser welding provided by the embodiments of the application is shown.

[0019] Explanation of reference signs: main stress transmission path excavation module 10, limiting strategy determination module 20, and welding driving module 30. DETAILED DESCRIPTION

[0020] In order to further illustrate the technical means and effects adopted by the application to achieve the predetermined invention purposes, the specific embodiments, structures, features and effects according to the application are described in detail as follows in combination with the drawings and the preferred embodiments.

[0021] The embodiments of the application provide an intelligent positioning processing method for a seat framework based on laser welding, as shown in Figure 1 The method comprises the following steps.

[0022] In step S100, the main stress transmission path of the seat framework is excavated based on a scene working condition, and a digital load atlas is constructed, wherein the digital load atlas is marked with pressure-bearing welds and tension-bearing welds.

[0023] Specifically, by computer aided engineering software, the mechanical response of the seat skeleton under the simulation of actual use load is analyzed by finite element method, and the main stress path is identified. Specifically, the material properties, constraint conditions and load conditions are set in the software, the statics or fatigue analysis solver is run, and the high stress area and force flow direction are extracted from the stress distribution cloud diagram in the calculation result. According to the relative relationship between stress direction and weld, the welds bearing compressive stress are marked as bearing welds, and the welds bearing tensile stress are marked as tension welds, and a digital model containing these marked information, namely digital load atlas, is generated.

[0024] For example, using ANSYS or Abaqus software, the three-dimensional model of the seat skeleton is imported, the pressure simulating the sitting posture of the human body and the inertial force during vehicle acceleration are applied for finite element calculation. The software outputs a stress cloud diagram. According to the cloud diagram, at the connection between the seat skeleton and the backrest skeleton, a main stress transmission path is identified, and the welds on the path are marked as blue in the three-dimensional model, representing bearing welds; at the hook point where the seat skeleton is connected with the chassis, a main tensile force transmission path is identified, and the welds at this point are marked as red, representing tension welds. All these marked information together constitutes a digital load atlas.

[0025] In one possible implementation, the digital load atlas is constructed, and step S100 further includes step S110 of constructing a lightweight assembly architecture of the seat skeleton according to the pre-processed part assembly and welding assembly requirements. Specifically, in the three-dimensional computer aided design software, each part to be welded, such as sheet metal stamping parts, pipe fittings and the like, is virtually assembled according to the final welding assembly form to form a simplified digital model containing only necessary geometric and assembly relationship information. This model removes unnecessary details such as fillets, chamfers, and clearly defines the contact relationship between parts and the position of the welding edge, which is used to provide a calculation basis for stress analysis.

[0026] For example, in CATIA or SolidWorks software, the three-dimensional models of the seat cushion skeleton, backrest skeleton, side plate, angle adjuster and other parts are imported into an assembly file. By defining face fitting, hole alignment and other constraints, these parts are positioned to their final welded positions. At the same time, the model is simplified, small features that do not affect the overall structural stiffness analysis are removed, and a lightweight assembly model is generated.

[0027] Step S120, in the lightweight assembly architecture, stress cycle marking is performed in the main stress transmission path, and the weld type marking constitutes the digital load map. Specifically, through the data interface of the software, the main stress transmission path result file obtained by the finite element software analysis is imported into the CAD software and overlaid on the assembly model. According to the stress size and direction, the weld lines in the model are color-coded or attribute-marked to distinguish between pressure-bearing welds and tension-bearing welds, and to mark stress concentration or fatigue danger areas.

[0028] For example, by software integration, the stress cloud data calculated by Abaqus is imported into the lightweight assembly in CATIA, the welds through which the high stress flows are identified, and different attributes are given to these weld line segments according to preset rules such as the angle between the stress direction and the weld normal. Finally, in the CATIA model, the pressure-bearing welds are displayed in one color and line type, and the tension-bearing welds are displayed in another color and line type, constituting a visual digital load map.

[0029] Step S200, according to the positioning agent, the digital load map is taken as the reference to make two-stage decisions of clamping limiting and clamping preloading, and the limiting strategy is determined.

[0030] Specifically, a trained intelligent decision system, i.e. a positioning agent, is used to plan how to clamp and pre-deform the seat skeleton based on the information of the digital load map in two steps, where the first step is clamping limiting and the second step is clamping preloading. The agent is a software module that embeds expert knowledge or learned rules about clamp design and welding deformation.

[0031] In one possible implementation, the construction of the positioning agent further includes step S210 of constructing a first positioning component based on clamping limiting and a second positioning component based on clamping preloading, cascading the first positioning component and the second positioning component to generate a positioning decision architecture. Specifically, the internal logical structure of the positioning agent is constructed in the software, the first positioning component is a rule base for determining the positions of the clamp positioning pins and support blocks according to the mechanical key points. The second positioning component is another rule base for calculating the preloading force and displacement required to be applied on the flexible clamp actuator. The two components are connected in sequence, the rigid positioning scheme is output by the first positioning component first, and then it is taken as the input of the second positioning component to calculate the flexible preloading scheme.

[0032] For example, the positioning decision architecture can be a rule-based expert system. The first positioning component contains a series of IF-THEN rules, such as “IF (the node is the starting point of the main stress path) THEN (set it as the main positioning reference, arrange the fixed supports according to the 3-2-1 positioning principle)”. The second positioning component contains another set of rules, such as “IF (the location node is near the tensile weld) THEN (apply a displacement command to the actuator at this point to pre-tension the material)”. The rule bases of the two components are integrated into a software framework.

[0033] At step S220, the digital load map is embedded in the positioning decision architecture, and supervised training is performed until convergence, to generate a positioning agent. Specifically, a large number of existing, successful seat skeleton welding fixture design schemes and their corresponding digital load maps are used as training data sets. The key path, location node, and weld type in the digital load map are used as input features, and the excellent fixture scheme designed by artificial experts, including the position of the limit point and the preload parameter, is used as the target label, to perform supervised learning on the positioning decision architecture. By iteratively adjusting the internal parameters of the model, the output of the model gradually approaches the expert scheme, until the prediction error stabilizes within an acceptable range.

[0034] For example, 100 sets of historical data are collected, each set including a digital load map of a seat skeleton and a fixture design drawing of the skeleton that has been verified through practice, wherein the fixture design drawing indicates the positions and parameters of all positioning pins, support blocks, and actuators. These data are input into a machine learning platform, and a neural network model is trained using the map data, so that the model learns to map from mechanical characteristics to the optimal fixture configuration. After training is complete, the model is the positioning agent.

[0035] In one possible implementation, the clamping limit decision is made, and step S200 further includes step S230 of positioning the key path and location node of the force flow transmission in the seat skeleton assembly according to the first positioning component and the digital load map. Specifically, the positioning agent reads the digital load map file, analyzes the high stress area and force flow arrow marked therein, identifies the key geometric positions where the force flow starts, ends, and turns, and extracts the coordinates of these positions as candidate positioning points.

[0036] For example, the positioning agent analyzes the digital load map and finds a high stress transmission band from the front end of the seat cushion to the angle adjuster. The positioning agent automatically identifies the starting point, end point, and intermediate stress concentration inflection point of the band, and records the three-dimensional coordinates of the three points.

[0037] Step S240, according to the key path and the position node, determine the primary and secondary positioning reference. Specifically, based on the positioning principle in mechanical manufacturing, such as the 3-2-1 principle, the key position nodes found in step S230 are sorted, and the points that limit most of the degrees of freedom of the part and play a core role in force flow transmission are selected as the primary positioning reference; the points that assist in limiting the remaining degrees of freedom or are located in the secondary force flow path are selected as the secondary positioning reference.

[0038] For example, the positioning agent determines the angle adjuster connecting hole that limits three translational and two rotational degrees of freedom as the primary positioning reference according to the rules; and determines the center of the front end beam of the cushion that limits the last rotational degree of freedom as the secondary positioning reference.

[0039] Step S250, for the primary and secondary positioning reference, the primary positioning reference is subjected to rigid limiting decision to determine the first clamping limit, and the secondary positioning reference is subjected to flexible limiting decision based on controllable degrees of freedom to determine the second clamping limit, wherein the second clamping limit is integrated with a precision actuator at each clamping action point, and the clamping limit result is determined according to the first clamping limit and the second clamping limit. Specifically, for the primary positioning reference, a non-movable and rigid positioning element, such as a fixed pin, a support block, is used to form the first clamping limit. For the secondary positioning reference, a flexible element that can be actively controlled, such as a servo electric cylinder, a pneumatic cylinder or a hydraulic cylinder, is used to form the second clamping limit. These actuators are integrated with displacement or force sensors and can control their extension and thrust. In this way, the stability of the primary positioning is ensured, and controllable pre-deformation is allowed at the secondary positioning point.

[0040] For example, the decision output is: at the primary positioning reference angle adjuster hole, a fixed cylindrical pin and a diamond pin are arranged. At the secondary positioning reference front end beam of the cushion, two top rods driven by servo motors are arranged, and the initial position, advancing speed and final stroke of the two top rods are programmable controlled.

[0041] In a possible implementation, the clamping pre-load application decision is made, and step S200 further includes step S260 of determining a pre-displacement field, wherein the pre-displacement field is a pre-strain applied before welding. Specifically, by welding simulation calculation, the shape change amount that needs to be actively generated by the framework before welding in order to offset the welding deformation and working load deformation, i.e. the pre-displacement field, is determined.

[0042] Step S270, according to the second positioning assembly, taking the second clamping limit as the clamping action point, pre-strain parameter analysis and iterative optimization based on the pre-displacement field are performed to determine the flexible limit parameter, wherein the flexible limit parameter at least includes the force direction and displacement stroke of each action point, and a clamping preload application result is generated according to the flexible limit parameter. Specifically, the second positioning assembly receives the global pre-displacement field and performs calculation with the flexible actuator position determined in step S250 as the actuator. The global pre-displacement field is decomposed into displacement instructions and directions that each flexible actuator needs to execute individually through finite element inverse operation, so as to determine the specific flexible limit parameter.

[0043] For example, the pre-displacement field requires the seat cushion front cross beam to generate a downward 1.5 mm and slightly twisted deformation. The second positioning assembly decomposes through calculation that the left actuator needs to move downward 1.8 mm and the right actuator needs to move downward 1.2 mm. These two displacement amounts and directions are the determined flexible limit parameters.

[0044] In a possible implementation, the step S260 of determining the pre-displacement field further includes step S261 of constructing a simulation field based on the clamping limit result to simulate and determine the ideal residual stress distribution after welding. Specifically, a virtual welding model is established using a welding simulation software with the fixture scheme determined in step S250 as the boundary condition. The software calculates the stress and deformation distribution remaining in the skeleton after welding by simulating the process of welding heat source movement, material melting, thermal expansion and cold contraction.

[0045] For example, in the SYSWELD software, the seat skeleton model is imported, the constraint positions of the fixed pins and the flexible top rods are set, and the power, speed and path of the laser welding are defined. The software runs the thermal-mechanical coupling analysis and finally outputs a cloud chart showing which areas of the skeleton after welding exist tensile stress and which areas exist compressive stress.

[0046] Step S262, according to the residual stress distribution, the first pre-strain to be applied before welding is reversely calculated. Specifically, the welding simulation software has an optimization function, and its algorithm is to iteratively adjust the initial displacement boundary conditions of the pre-welding model to reduce the final deformation. Technically, the deformation field obtained by welding simulation is taken as the opposite or the best pre-deformation amount that can make the final deformation tend to zero is calculated through the adjoint variable method, and this amount is the first pre-strain.

[0047] For example, the simulation shows that the seat cushion edge after welding will be upwarping 2 mm. Then, through the optimization module of the software, it is automatically calculated that if the seat cushion edge is pressed downward 1.5 mm by the fixture before welding, that is, a negative pre-displacement is applied, the upwarping amount after welding will be reduced to less than 0.5 mm. This downward pressing 1.5 mm instruction is the first pre-strain.

[0048] Step S263, scanning the digital load map to determine the second pre-strain based on the working load, coupling the first pre-strain and the second pre-strain to generate the pre-displacement field. Specifically, the deformation of the framework under the working load is read from the digital load map. The first pre-strain for offsetting the welding deformation and the second pre-strain for optimizing the stress distribution under the working condition are vector superimposed. This superimposed and comprehensive displacement instruction set is the final pre-displacement field to be implemented by the flexible clamp.

[0049] For example, the digital load map shows that the middle of the seat cushion will sag when subjected to a load of 100 kg. In order to improve the rigidity of the seat, a pre-arching upward of the framework, i.e. the second pre-strain, can be provided. The downward pressure required for welding anti-deformation and the pre-arching under the working condition are combined in space to generate a non-uniform pre-displacement field, which is issued to each flexible actuator.

[0050] Step S300, according to the limiting strategy, the parts of the seat framework are assembled and clamped, the three-dimensional coordinates of the welds are scanned and converted, the welding agent is combined to determine the welding driving parameters, and the automatic welding driving of the laser welding assembly is executed.

[0051] Specifically, the parts are clamped using the clamp system manufactured according to the positioning agent decision, then the accurate position of the actual clamped weld is obtained by using a 3D scanner, and finally the welding agent is called to calculate the welding process parameters to drive the robot to complete the welding.

[0052] In a possible implementation, the automatic welding driving of the laser welding assembly is executed, and step S300 further includes step S310, according to the clamping limiting result and the clamping preloading application result, the parts of the seat framework are assembled and clamped, and multi-angle high-speed scanning after clamping is performed to convert the three-dimensional coordinates of the welds of the laser welding assembly. Specifically, each seat framework part is placed on the clamp, and the rigid positioning part and the flexible actuator of the clamp fix and pre-strain it to the position. A 3D vision sensor installed on the production line, such as a laser scanner or a structured light camera, is used to quickly scan the clamped assembly from multiple angles to obtain point cloud data with weld features. Through coordinate transformation, the weld line features in these point clouds are converted to three-dimensional coordinates in the robot base coordinate system.

[0053] For example, a laser displacement sensor group of Keyence is used to scan the clamped framework above the clamp to generate a point cloud of millions of points. The system software extracts all the continuous three-dimensional space curves of the lap joints and corner seams to be welded from the point cloud through a feature recognition algorithm, and the coordinates of these curves are the paths that the robot needs to follow.

[0054] Step S320, constructing a state space between the strength and topography gap of the weld, constructing an action space between the weld trajectory, power, and speed, constructing a welding agent based on the linear relationship between the action space and the state space, wherein the welding agent is an embedded plug-in of the laser welding assembly. Specifically, the state space refers to a set of input variables that affect the welding quality, which is the joint form of the weld and the assembly gap between the parts. The joint form of the weld includes lap joint, corner joint, etc. Different joint forms have different inherent strength characteristics and required welding processes. The action space refers to a set of controllable welding process parameters, which is the movement trajectory of the laser head, the output power of the laser, and the welding speed. The welding agent is a mathematical model, which can be a regression model, a lookup table, or a neural network, and establishes a mapping relationship from state to action.

[0055] For example, the welding agent can be an experience database, i.e. a lookup table. This table is established on the basis of a large number of process tests, and its structure is as follows: if the state is a lap joint and the gap is 0.1 mm, then the corresponding action is a laser power of 3 kW, a welding speed of 2 m / min, and a circular beam oscillation trajectory.

[0056] Step S330, establishing communication interaction between the welding agent and the mechanical arm and the end effector, and driving the laser welding assembly to perform automatic positioning welding. Specifically, the industrial field bus or the robot special interface is used to connect the industrial computer / PLC containing the welding agent to the robot controller and the laser power supply. The welding agent sends the calculated weld coordinates and welding parameters to the robot controller, which drives the mechanical arm to move the laser welding gun according to the trajectory, and simultaneously controls the laser to emit light according to the set parameters.

[0057] For example, in the upper computer, the welding agent software packs the weld three-dimensional coordinate points obtained in step S310 and the welding parameters queried in step S320 into an instruction sequence through the EtherCAT bus, and sends it to the controller of the robot and the numerical control system of the laser. The robot and the laser cooperate to complete the automatic welding.

[0058] In one possible implementation, after the laser welding assembly performs automatic welding driving, the method further includes step S400, adding an acoustic emitter to the end effector, wherein the acoustic emitter has a first time delay based on the welding operation. Specifically, an acoustic emission sensor is installed beside the laser welding head. The acoustic emission sensor is set to trigger a delay so that it starts listening a little later after the laser starts welding, i.e. the first time delay is to avoid the strong initial noise generated when the laser hits the material, and only collect the acoustic emission signals after the weld pool is formed.

[0059] For example, a piezoelectric ceramic acoustic emission sensor is attached to the welding torch by a magnetic base. A time delay relay is set in the control system. When the welding start signal is received, the power supply of the acoustic emission acquisition system is turned on after 50 milliseconds, and the recording begins.

[0060] In step S500, the acoustic emission sensor detects the welding defects during the laser welding process, and analyzes the welding defects through the acoustic emission signals, wherein the welding defects at least include pores and incomplete fusion defects. Specifically, the acoustic emission sensor collects the stress wave signals generated by the plastic deformation, crack generation, gas escape and other events in the material during the welding process in real time. After the signals are digitized by the acquisition card, they are sent to the signal processing unit. The unit analyzes the amplitude, frequency, energy and count of the signals, and performs pattern matching with the acoustic emission fingerprint database of known defects to determine whether there are pores, incomplete fusion and other defects in real time.

[0061] For example, the acoustic emission acquisition card records the signals at a sampling rate of 1 megahertz per second. The software calculates the effective value of the signals in real time. When a series of high-frequency, low-amplitude burst signals are detected in the signals, they are identified as the characteristics of pore formation; when a persistent high-energy signal is detected, it is identified as incomplete fusion.

[0062] In step S600, the positioning based on the welding defects and the rework welding driving are performed. Specifically, the defect processing decision is performed. Once the defect is identified in step S500, the system triggers a rework process.

[0063] In one possible implementation, step S600 further includes step S610. The laser welding assembly is a multi-axis assembly. When there is a welding defect, a first collaborative robot is driven to determine the rework positioning result through the positioning of the welding defect and the three-dimensional coordinate conversion based on the first collaborative robot. Specifically, the main welding robot detects the defect through the acoustic emission sensor integrated on the welding torch during welding. The system records the three-dimensional coordinates of the tool center point of the main welding robot in its own coordinate system at this time. In the pre-calibrated automatic work station, the position relationship between the main robot and the base of the first collaborative robot is known. Through a fixed base coordinate system conversion matrix, the coordinates reported by the main robot are initially converted to the base coordinate system of the first collaborative robot to obtain a rough target coordinate. The first collaborative robot moves to the vicinity of the rough target coordinate point, and the vision sensor carried at the end of the first collaborative robot performs local high-definition scanning on the weld. The exact profile and center of the pores are identified through the image processing algorithm. Through the conversion relationship between the vision sensor and the end of the mechanical arm which is pre-calibrated, the pixel position of the defect in the image is converted to the three-dimensional space coordinates that the end of the first collaborative robot needs to reach, i.e. the rework positioning result.

[0064] Step S620, according to the rework positioning result, drive the first collaborative mechanical arm, perform the rework welding of the welding defect. Specifically, the control system automatically calls the preset repair welding parameters from the process parameter library according to the defect type. These parameters are different from the main welding, which adopts lower laser power, slower welding speed and increased laser beam swing to avoid overburning and better fill the defects. The motion controller of the first collaborative mechanical arm receives the accurate coordinates calculated in step S610. The controller plans a safe and efficient path from the current point to the accurate coordinates. The mechanical arm moves and accurately positions the end welding gun above the defect. After the mechanical arm is in place, the system sends a start command to the laser integrated into the first collaborative mechanical arm through the field bus, and sets the required power, frequency and other parameters for repair. The first collaborative mechanical arm performs repair welding at the accurate coordinate point to make the metal re-solidify to eliminate the defect. After completion, the laser is turned off and the mechanical arm is moved away.

[0065] The embodiment of the present application adopts the method of excavating the main stress transmission path of the seat framework to construct a digital load map marked with pressure-bearing and tension-bearing welds, determines a limiting strategy according to the positioning intelligent agent based on the map, clamps and limits according to the limiting strategy, determines the three-dimensional coordinates of the welds through scanning conversion, determines the parameters in combination with the welding intelligent agent, and drives the laser welding assembly to automatically weld, thereby solving the technical problem that the fixture positioning and the welding process are independent of each other in the existing laser welding of the seat framework, the clamping point and the key point of the force flow are misaligned, and the welding precision and the structural reliability are affected, realizing the collaborative optimization of the clamping and limiting and the welding driving, and further improving the welding precision and the structural reliability.

[0066] In the foregoing, reference is made to Figure 1 The laser welding-based intelligent positioning processing method for the seat framework according to the embodiment of the present application is described in detail. Next, the laser welding-based intelligent positioning processing system for the seat framework according to the embodiment of the present application will be described with reference to Figure 2 The laser welding-based intelligent positioning processing system for the seat framework according to the embodiment of the present application is described in detail. Next, the laser welding-based intelligent positioning processing system for the seat framework according to the embodiment of the present application will be described with reference to

[0067] The laser welding-based intelligent positioning processing system for the seat framework according to the embodiment of the present application is used to solve the technical problem that the fixture positioning and the welding process are independent of each other in the existing laser welding of the seat framework, the clamping point and the key point of the force flow are misaligned, and the welding precision and the structural reliability are affected, realize the collaborative optimization of the clamping and limiting and the welding driving, and further improve the welding precision and the structural reliability. The laser welding-based intelligent positioning processing system for the seat framework includes a main stress transmission path excavation module 10, a limiting strategy determination module 20 and a welding driving module 30.

[0068] The main stress transmission path mining module 10 is used for mining the main stress transmission path of the seat framework based on a scene working condition, and constructs a digital load atlas, wherein the digital load atlas is marked with a pressure-bearing weld and a tension-bearing weld; the limiting strategy determination module 20 is used for determining a limiting strategy by means of a positioning intelligent agent and taking the digital load atlas as a reference to make two-stage decisions of clamping limiting and clamping preloading; and the welding driving module 30 is used for performing assembly clamping limiting of parts of the seat framework according to the limiting strategy, scanning and converting to determine three-dimensional coordinates of a weld, combining a welding intelligent agent to determine welding driving parameters, and performing automatic welding driving of a laser welding assembly.

[0069] The main stress transmission path mining module 10 is used for mining the main stress transmission path of the seat framework based on a scene working condition, and constructs a digital load atlas, wherein the digital load atlas is marked with a pressure-bearing weld and a tension-bearing weld; the limiting strategy determination module 20 is used for determining a limiting strategy by means of a positioning intelligent agent and taking the digital load atlas as a reference to make two-stage decisions of clamping limiting and clamping preloading; and the welding driving module 30 is used for performing assembly clamping limiting of parts of the seat framework according to the limiting strategy, scanning and converting to determine three-dimensional coordinates of a weld, combining a welding intelligent agent to determine welding driving parameters, and performing automatic welding driving of a laser welding assembly.

[0070] The main stress transmission path mining module 10 is used for mining the main stress transmission path of the seat framework based on a scene working condition, and constructs a digital load atlas, wherein the digital load atlas is marked with a pressure-bearing weld and a tension-bearing weld; the limiting strategy determination module 20 is used for determining a limiting strategy by means of a positioning intelligent agent and taking the digital load atlas as a reference to make two-stage decisions of clamping limiting and clamping preloading; and the welding driving module 30 is used for performing assembly clamping limiting of parts of the seat framework according to the limiting strategy, scanning and converting to determine three-dimensional coordinates of a weld, combining a welding intelligent agent to determine welding driving parameters, and performing automatic welding driving of a laser welding assembly.

[0071] The main stress transmission path mining module 10 is used for mining the main stress transmission path of the seat framework based on a scene working condition, and constructs a digital load atlas, wherein the digital load atlas is marked with a pressure-bearing weld and a tension-bearing weld; the limiting strategy determination module 20 is used for determining a limiting strategy by means of a positioning intelligent agent and taking the digital load atlas as a reference to make two-stage decisions of clamping limiting and clamping preloading; and the welding driving module 30 is used for performing assembly clamping limiting of parts of the seat framework according to the limiting strategy, scanning and converting to determine three-dimensional coordinates of a weld, combining a welding intelligent agent to determine welding driving parameters, and performing automatic welding driving of a laser welding assembly.

[0072] The clamping preload application decision is made, and the limiting strategy determination module 20 can further include: a pre-displacement field determination unit for determining a pre-displacement field, wherein the pre-displacement field is a pre-strain applied before welding; a flexible limiting parameter determination unit for determining a flexible limiting parameter based on the pre-displacement field with the second clamping limiting as the clamping action point of the second positioning assembly, wherein the flexible limiting parameter at least includes the force direction and displacement stroke of each action point, and a clamping preload application result is generated according to the flexible limiting parameter.

[0073] The pre-displacement field is determined, and the pre-displacement field determination unit can further include: an analog field construction subunit for constructing an analog field based on the clamping limiting result, simulating and determining the ideal residual stress distribution after welding; a first pre-strain calculation subunit for inversely calculating the first pre-strain to be applied before welding according to the residual stress distribution; a second pre-strain determination subunit for scanning the digital load spectrum to determine the second pre-strain based on the working load; and a pre-displacement field generation subunit for coupling the first pre-strain and the second pre-strain to generate the pre-displacement field.

[0074] The specific configuration of the welding driving module 30 is described in detail as follows: as described above, the laser welding assembly is automatically driven for welding, and the welding driving module 30 can further include: a multi-angle high-speed scanning unit for assembling and clamping the parts of the seat framework according to the clamping limiting result and the clamping preload application result, and performing multi-angle high-speed scanning after clamping to convert into the weld three-dimensional coordinates of the laser welding assembly; a welding agent construction unit for constructing a state space based on the weld strength and topography gap, constructing an action space based on the weld trajectory, power and speed, and constructing a welding agent based on the linear relationship between the action space and the state space, wherein the welding agent is an embedded plug-in of the laser welding assembly; and an automatic positioning welding unit for establishing communication and interaction between the welding agent and the mechanical arm and the end effector to drive the laser welding assembly for automatic positioning welding.

[0075] After the laser welding assembly is automatically driven for welding, the system can further include: a sound emitter installation module for installing a sound emitter on the end effector, wherein the sound emitter has a first time delay based on the welding operation; a welding defect analysis module for performing welding energy efficiency detection by the sound emitter along with the welding process of the laser welding assembly, and analyzing the welding defects through the sound emission signal, wherein the welding defects at least include pores and incomplete fusion defects; and a rework welding driving module for performing positioning and rework welding driving based on the welding defects.

[0076] The rework welding driving module can further include: a rework positioning result determination unit configured to, when there is a welding defect, drive the first collaborative robot arm, determine a rework positioning result by positioning the welding defect and three-dimensional coordinate conversion based on the first collaborative robot arm, when the laser welding assembly is a multi-axis assembly.

[0077] The seat framework intelligent positioning processing system based on laser welding provided by the embodiment of the present application can execute the seat framework intelligent positioning processing method based on laser welding provided by any embodiment of the present application, and has the corresponding function modules and beneficial effects of the execution method.

[0078] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, however, any number of different modules can be used and run on the user terminal and / or server, and each unit and module included is only divided according to the functional logic, but is not limited to the above division, as long as the corresponding functions can be realized; in addition, the specific names of each functional unit are only for easy mutual differentiation, and do not limit the protection scope of the present application.

[0079] The above is only a preferred embodiment of the present application, and does not limit the present application in any form. Although the present application has been disclosed as above with the preferred embodiment, it is not intended to limit the present application. Any person skilled in the art can make some changes or modifications to the above disclosed technical content without departing from the scope of the technical solution of the present application, and any modification, equivalent change and modification made to the above embodiment according to the technical essence of the present application still belong to the scope of the technical solution of the present application.

Claims

1. A method for intelligent positioning and processing of seat frames based on laser welding, characterized in that, The method includes: The principal stress transmission path of the seat frame is mined based on the scenario working condition, and a digital load map is constructed, wherein the digital load map identifies pressure welds and tension welds. Based on the positioning agent and using the digital load map as a reference, a two-order decision is made on clamping limit and clamping preload application to determine the limit strategy; According to the limiting strategy, the seat frame is clamped and limited in assembly, the three-dimensional coordinates of the weld are scanned and converted to determine the weld seam, the welding driving parameters are determined in combination with the welding intelligent body, and the laser welding assembly is automatically driven for welding. The construction of the localization agent includes: Construct a first positioning component based on clamping limit and a second positioning component based on clamping preload, and cascade the first positioning component and the second positioning component to generate a positioning decision architecture; The digital payload map is embedded in the localization decision architecture and supervised to convergence to generate a localization agent; The clamping limit decision includes: Based on the first positioning component and the digital load map, locate the critical path and position node of force flow transmission in the seat frame assembly; Based on the critical path and location nodes, determine the primary and secondary positioning references; For the primary and secondary positioning references, a rigid limit application decision is made on the primary positioning reference to determine the first clamping limit, and a flexible limit application decision based on controllable degrees of freedom is made on the secondary positioning reference to determine the second clamping limit. Among them, each clamping action point of the second clamping limit integrates a precision actuator. The clamping limit result is determined based on the first clamping limit and the second clamping limit.

2. The intelligent positioning and processing method for seat frames based on laser welding as described in claim 1, characterized in that, Constructing a digital payload map includes: Based on the pre-processed parts and components and welding assembly requirements, a lightweight assembly structure for the seat frame is constructed. The digital load map is constructed by marking stress cycles and weld type in the lightweight assembly architecture according to the principal stress transmission path.

3. The intelligent positioning and processing method for seat frames based on laser welding as described in claim 1, characterized in that, Making clamping preload application decisions includes: Determine the pre-displacement field, wherein the pre-displacement field is the pre-strain applied before welding; Based on the second positioning component, with the second clamping limit as the clamping action point, pre-strain parameter analysis and iterative optimization based on the pre-displacement field are performed to determine the flexible limit parameters, wherein the flexible limit parameters include at least the force direction and displacement stroke of each action point; Based on the flexible limiting parameters, the clamping preload application result is generated.

4. The intelligent positioning and processing method for seat frames based on laser welding as described in claim 3, characterized in that, Determine the pre-displacement field, including: A simulation field is constructed based on the clamping and limiting results to simulate and determine the residual stress distribution after ideal welding. Based on the residual stress distribution, the first pre-strain to be applied before welding is calculated in reverse. Scan the digital load spectrum to determine the second pre-strain based on the working load; The first pre-strain and the second pre-strain are coupled to generate the pre-displacement field.

5. The intelligent positioning and processing method for seat frames based on laser welding as described in claim 4, characterized in that, Automated welding drive for laser welding assemblies includes: Based on the clamping limit result and the clamping preload application result, the seat frame is subjected to assembly-type clamping limit of parts, and multi-angle high-speed scanning is performed after clamping to convert it into three-dimensional coordinates of the weld seam of the laser welding assembly. A state space is constructed using weld strength and morphological gap, an action space is constructed using weld trajectory, power, and speed, and a welding intelligent agent is constructed using the linear relationship between the action space and the state space. The welding intelligent agent is an embedded plug-in of the laser welding assembly. Establish communication and interaction between the welding intelligent agent and the robotic arm and end effector to drive the laser welding assembly to perform automatic positioning welding.

6. The intelligent positioning and processing method for seat frames based on laser welding as described in claim 1, characterized in that, After performing automated welding drive on the laser welding assembly, the following is included: An acoustic emitter is added to the end effector, wherein the acoustic emitter has a first time delay based on the welding operation; As the laser welding assembly progresses, the acoustic emitter performs welding energy efficiency detection and analyzes welding defects through acoustic emission signals. The welding defects include at least porosity and lack of fusion. Perform welding based on the location and rework of welding defects.

7. The intelligent positioning and processing method for seat frames based on laser welding as described in claim 6, characterized in that, The laser welding assembly is a multi-axis assembly that drives the first collaborative robotic arm when welding defects exist. The resumption positioning result is determined by locating the welding defects and transforming the three-dimensional coordinates based on the first collaborative robotic arm; Based on the resumption of work positioning results, the first collaborative robotic arm is driven to perform resumption of welding of welding defects.

8. A seat frame intelligent positioning and processing system based on laser welding, characterized in that, The system is used to implement the intelligent positioning and processing method for seat frames based on laser welding as described in any one of claims 1-7, and the system comprises: The principal stress transmission path mining module is used to mine the principal stress transmission path of the seat frame based on the scenario working conditions and construct a digital load map, wherein the digital load map identifies pressure welds and tension welds. The limit strategy determination module is used to determine the limit strategy by making two-order decisions on clamping limit and clamping preload application based on the positioning agent and the digital load map. The welding drive module is used to perform assembly-type clamping and positioning of the seat frame parts according to the limiting strategy, scan and convert to determine the three-dimensional coordinates of the weld, combine with the welding intelligent agent to determine the welding drive parameters, and perform automated welding drive on the laser welding assembly.

Images