Welding robot group intelligent control method and system based on Internet of Things

Abstract

The invention discloses a welding robot group intelligent control method and system based on the Internet of Things, and belongs to the technical field of welding robots. The method comprises the following steps: setting a horizontally staggered and symmetrical station sequence on two sides of an assembly line based on welding process logic, and constructing a dynamic communication network consisting of a longitudinal feedback Internet of Things chain and a transverse feedback Internet of Things branch chain; the welding robot is endowed with a transverse cooperative capability (symmetric stations execute the same working mode) and a longitudinal cooperative capability (the whole-side robot is dynamically adjusted in parameters) through an internet-of-things chip; a welding working mode defined by the welding current and the wire feeding speed is pre-configured, and a wooden barrel dynamic model representation collaborative network with a longitudinal chain as the bottom face and a transverse branch chain as the top face is constructed; and a dynamic balance value is generated by calculating the transverse entropy difference and the longitudinal entropy flow, and welding parameters are optimized under the fusion depth and fusion width process constraint. Through the Internet of Things and the swarm intelligence technology, the collaborative efficiency of the welding robot and the welding quality stability are remarkably improved, and the process error is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of welding robots, and specifically to a swarm intelligence control method and system for welding robot groups based on the Internet of Things. Background Art

[0002] In the field of welding automation, multi-robot collaborative control is a key technology to improve the welding quality and efficiency of complex workpieces; traditional welding robot control methods usually rely on single robot independent operation or simple master-slave control, lacking effective horizontal collaboration (synchronization of symmetric station parameters) and vertical collaboration (optimization of parameter transfer between stations) mechanisms, resulting in the following problems:

[0003] Low collaboration efficiency: The asynchronous operation caused by the horizontal misalignment layout of welding robots on both sides of the assembly line is likely to cause welding seam splicing deviation, increasing the post-processing cost;

[0004] Poor process adaptability: The adjustment of parameters such as welding current and wire feeding speed depends on manual experience, and it is difficult to respond to material changes (such as the difference in thermal conductivity between stainless steel and aluminum alloy) and complex process constraints (such as penetration and bead width tolerances) in real time;

[0005] Insufficient quality stability: Lack of dynamic collaborative modeling and feedback optimization, parameter fluctuations during the welding process are likely to cause defects such as lack of fusion and weld cracking, especially with a high defect rate during multi-robot segmented welding.

[0006] In the prior art, the application of the Internet of Things in industrial control is mostly limited to equipment status monitoring, and no systematic solution for welding robot collaborative control has been formed. Therefore, there is an urgent need for a swarm intelligence control method combined with Internet of Things technology to solve the problems of parameter synchronization, process adaptation, and quality control during multi-robot welding by constructing a vertical and horizontal collaborative network and a dynamic optimization model. Summary of the Invention

[0007] The purpose of the present invention is to provide a swarm intelligence control method and system for welding robot groups based on the Internet of Things to solve the problems raised in the above background art.

[0008] To solve the above technical problems, the present invention provides the following technical solutions:

[0009] A swarm intelligence control system for welding robot groups based on the Internet of Things, the system includes: a station layout and network construction module, a collaborative ability assignment module, a working mode configuration and model construction module, and a collaborative control and parameter optimization module;

[0010] The station layout and network construction module sets the station sequence on both sides of the assembly line based on the welding process logic, forms a three-dimensional operation parallel group with horizontally misaligned and symmetric welding robots, and establishes a vertical feedback Internet of Things chain and a horizontal feedback Internet of Things branch chain;

[0011] The collaborative ability endowment module endows the welding robot with horizontal collaborative ability and vertical collaborative ability through the horizontal feedback IoT branch chain and the vertical feedback IoT chain;

[0012] The working mode configuration and model construction module is used to pre-configure the welding working mode and construct a wooden barrel dynamic model to represent the robot collaborative network;

[0013] The collaborative control and parameter optimization module is used to establish a swarm intelligence control collaborative model and a swarm intelligence control feedback model, and optimize the welding current and wire feeding speed in real time to meet the process constraints of penetration and bead width.

[0014] Further, the station layout and network construction module includes an operation parallel group construction unit and a feedback network setting unit;

[0015] The operation parallel group construction unit is used to form an operation parallel group with two welding robots that are horizontally misaligned and symmetric on both sides of the assembly line to perform segmented welding of the same welding object area;

[0016] The feedback network setting unit is used to set a vertical feedback IoT chain between the welding robots on one side of the assembly line based on the station sequence, and set a horizontal feedback IoT branch chain between the symmetric robots in the operation parallel group.

[0017] Further, the collaborative ability endowment module includes a horizontal collaborative control unit and a vertical collaborative control unit;

[0018] The horizontal collaborative control unit enables the symmetric welding robots in the operation parallel group to execute the same welding working mode through the horizontal feedback IoT branch chain and the IoT chip;

[0019] The vertical collaborative control unit dynamically adjusts the welding working mode of all the welding robots on one side through the vertical feedback IoT chain and the IoT chip.

[0020] Further, the working mode configuration and model construction module includes a welding working mode pre-configuration unit and a wooden barrel dynamic model construction unit;

[0021] The welding working mode pre-configuration unit is used to define the welding working mode composed of the welding current and the wire feeding speed;

[0022] The wooden barrel dynamic model construction unit uses the vertical feedback IoT chain as the bottom surface of the wooden barrel dynamic model and the horizontal feedback IoT branch chain as the top surface of the wooden barrel dynamic model, and symmetrically connects the bottom surface and the top surface through the operation parallel group to form a wooden barrel dynamic model representing the robot collaborative network.

[0023] Further, the collaborative control and parameter optimization module includes a collaborative model calculation unit and a feedback optimization unit;

[0024] The collaborative model calculation unit is used to record the real-time operation state vectors of the welding robots at each station, and calculate the horizontal entropy difference and the vertical entropy flow;

[0025] The feedback optimization unit is used to map the horizontal entropy difference and the vertical entropy flow to the wooden barrel dynamic model, generate a dynamic balance value, and minimize the balance value under the constraints of the penetration tolerance and the bead width tolerance, where the penetration tolerance and the bead width tolerance are determined according to the maximum and minimum values of the penetration and the bead width required by the welding process.

[0026] An Internet of Things-based intelligent control method for a group of welding robots, this method includes the following steps:

[0027] Step S1: Set the station order on both sides of the production line based on the welding process logic, form a three-dimensional operation parallel group with two horizontally misaligned and symmetric welding robots, and establish a dynamic communication connection between the robots through the longitudinal feedback Internet of Things chain and the horizontal feedback Internet of Things branch chain;

[0028] Step S2: Based on the horizontal feedback Internet of Things branch chain, endow two horizontally misaligned and symmetric welding robots with horizontal collaborative capabilities so that they execute the same welding working mode; based on the longitudinal feedback Internet of Things chain, endow the welding robots on the same side with longitudinal collaborative capabilities to dynamically adjust the welding parameters of the entire side of the robots;

[0029] Step S3: Pre-configure multiple groups of welding working modes, which are defined by the welding current and the wire feeding speed, and construct a wooden barrel dynamic model to characterize the robot collaborative network, where the bottom surface is the longitudinal feedback Internet of Things chain, the top surface is the horizontal feedback Internet of Things branch chain, and the bottom surface and the top surface are connected through the symmetry of the operation parallel group;

[0030] Step S4: Establish a group intelligent control collaborative model, collect the operation state parameters of each welding robot in real time, calculate the horizontal entropy difference and the vertical entropy flow, map them to the wooden barrel dynamic model to generate a dynamic balance value, and minimize the balance value by optimizing the welding current and the wire feeding speed under the process constraint conditions of meeting the penetration and the bead width.

[0031] Further, the specific implementation process of the step S1 includes:

[0032] Step S11: Set the station order according to the welding process logic, the stations have horizontal misalignment symmetry on both sides of the production line, form an operation parallel group with two horizontally misaligned and symmetric welding robots, and perform segmented welding of the same welding object area at the stations;

[0033] Step S12: Select all the welding robots on one side of both sides of the assembly line, set up a longitudinal feedback Internet of Things chain for the welding robots based on the station sequence, and set up a transverse feedback Internet of Things branch chain between two welding robots with horizontal misalignment symmetry;

[0034] In the above method, through the horizontal misalignment symmetric station layout and operation parallel group division, the segmented collaborative operation of the same welding area is realized, improving the welding coverage efficiency and the seam splicing accuracy; the setting of the longitudinal feedback Internet of Things chain and the transverse feedback Internet of Things branch chain constructs a three-dimensional communication network of "longitudinal sequential transmission - transverse symmetric coupling", providing a hardware basis for real-time data interaction and collaborative control between robots, and avoiding the delay and synchronization error of traditional single-chain communication.

[0035] Further, the specific implementation process of step S2 includes:

[0036] Step S21: Based on the transverse feedback Internet of Things branch chain, use the Internet of Things chip to enable two welding robots with a transverse feedback Internet of Things branch chain relationship to have transverse collaborative capabilities, and the transverse collaborative capabilities are used to enable two welding robots with horizontal misalignment symmetry to execute the same welding work mode;

[0037] Step S22: Based on the longitudinal feedback Internet of Things chain, use the Internet of Things chip to enable the welding robots on one side with a station sequence relationship to have longitudinal collaborative capabilities, and the longitudinal collaborative capabilities are used to adjust the welding work modes of all the welding robots on one side of both sides of the assembly line;

[0038] In the above method, transverse collaboration: ensuring that the welding robots with horizontal misalignment symmetry execute the same work mode is beneficial to eliminating the heat input differences caused by the station layout (such as inconsistent welding current and wire feeding speed), beneficial to avoiding the stress concentration defects caused by asymmetric penetration depth and width on both sides of the weld, and beneficial to improving the mechanical property consistency of the welded joint;

[0039] Longitudinal collaboration: Dynamically adjusting the welding parameters of all the robots on one side through the longitudinal chain enables the process parameter fluctuations in the previous station to be sensed and compensated in real time by the subsequent station, forming a parameter self-calibration mechanism at the assembly line level, which is beneficial to adapting to the position deviation or material thickness change during the workpiece transmission process.

[0040] Further, the specific implementation process of step S3 includes:

[0041] Step S31: Preset several welding work modes distinguished according to welding process requirements, and the welding work modes are configured by welding current and wire feeding speed, and denote any i-th welding work mode as M i ={I i , SV i}, where I i and SV i respectively represent the welding current and wire feeding speed in the i-th welding working mode;

[0042] Step S32: Construct a barrel dynamic model to characterize the dynamic cooperation network among welding robots. The bottom surface of the barrel dynamic model is a longitudinal feedback Internet of Things chain, and the top surface of the barrel dynamic model is a transverse feedback Internet of Things branch chain. The connection relationship between the bottom surface and the top surface is the symmetry relationship of the operation parallel group on both sides of the production line. Then, for the welding robot U on one side of the x-th station x , there is a connection edge F in the barrel dynamic model x ={U x →U x+1 , U x →U x‘}, where U x →U x+1 exists in the longitudinal feedback Internet of Things chain, and U x →U x‘ exists in the transverse feedback Internet of Things branch chain. U x ‘ is the welding robot on the other side of the x-th station in the operation parallel group;

[0043] In the above method, the working mode is pre-configured: decouple the welding current and wire feeding speed into configurable working modes, support quick switching between different welding processes (such as MIG welding, TIG welding), meet the production requirements of multi-variety mixed lines, and provide standardized input parameters for the subsequent cooperation model;

[0044] Barrel dynamic model: Construct a cooperation network model with the longitudinal chain as the bottom surface and the transverse branch chain as the top surface, intuitively characterize the dynamic connection relationship among robots (such as the station sequence edge and the symmetry edge), provide a geometric modeling tool for quantifying the cooperation efficiency (such as the transverse entropy difference, longitudinal entropy flow), and facilitate the realization of global cooperation optimization by optimizing the model volume (dynamic balance value).

[0045] Further, the specific implementation process of step S4 includes:

[0046] Step S41: Construct a swarm intelligence control cooperation model, including a transverse cooperation model and a longitudinal cooperation model:

[0047] Based on the working mode of the welding machine, record the operation state vector of the welding robot U on one side of the x-th station x and the operation state vector of the welding robot U on the other side where t is the independent variable of the time node x‘ and and and The welding robot U operating in the i-th welding mode at the time node independent variable t respectively x The real-time welding current and wire feeding speed, and The welding robot U operating in the i-th welding mode at the time node independent variable t respectively x‘ The real-time welding current and wire feeding speed;

[0048] Lateral cooperation model:

[0049]

[0050] Longitudinal cooperation model:

[0051]

[0052] In the formula, H h (x) represents the lateral entropy difference, H z (x) represents the longitudinal entropy flow, t 1 is the end time within a sampling period, t 0 is the start time within a sampling period,

[0053] Step S42: Based on the swarm intelligence control cooperation model, construct a swarm intelligence control feedback model. Take the lateral entropy difference H h (x) as the side length of the connecting edge U x →U x‘ , take the longitudinal entropy flow H z (x) as the side length of the connecting edge U x →U x+1 , and map it to the wooden barrel dynamic model to generate the volume of the wooden barrel dynamic model as the wooden barrel dynamic balance value. The volume function is denoted as E;

[0054] Adjust the welding current and wire feeding speed through welding process constraint conditions to minimize the wooden barrel dynamic balance value:

[0055]

[0056] In the formula, D(x, t) is the empirical penetration depth linear function, and K D is the penetration depth coefficient, W(x, t) is the empirical penetration width linear function, and K W is the penetration width coefficient, b D and b W are both overshoot control coefficients; D 0 is the target penetration depth, and W 0 is the target penetration width, and D max and D min are the maximum and minimum penetration depths required by the welding process, respectively, and W max and W min are the maximum and minimum weld widths required by the welding process, respectively; δ D is the penetration tolerance, and δ W is the weld width tolerance, and

[0057] In the above method, the penetration coefficient varies with the thermal conductivity of the material. For example, for stainless steel K D = 0.08 mm / A, and for aluminum alloy K D = 0.05 mm / A; the weld width coefficient reflects the influence of the deposition rate on the weld width; the smaller the dynamic balance value of the wooden barrel, the smaller the error between all welding robots and the higher the collaborative efficiency;

[0058] Collaborative model calculation: Dynamically collect the real-time data of the welding current and wire feeding speed, quantify the collaborative state through the horizontal entropy difference (consistency of symmetric station parameters) and the vertical entropy flow (parameter transfer strength between stations), and convert the abstract collaborative ability into measurable physical indicators (such as the smaller the entropy value, the better the collaboration), providing a data basis for feedback control;

[0059] Feedback optimization: Minimize the dynamic balance value of the wooden barrel under the process constraints of penetration and weld width to ensure that the dynamic matching of the welding heat input (current) and the deposition rate (wire feeding speed) meets the requirements of the weld geometry size (such as penetration tolerance and weld width tolerance), suppressing defects such as lack of fusion and burn-through from the process source and improving the stability of welding quality.

[0060] Compared with the prior art, the beneficial effects achieved by the present invention are:

[0061] In a method and system for group intelligent control of welding robots based on the Internet of Things provided by the present invention, the station order of horizontal misalignment symmetry on both sides of the assembly line is set based on the welding process logic, and a dynamic communication network composed of a longitudinal feedback Internet of Things chain and a horizontal feedback Internet of Things branch chain is constructed; the welding robots are given horizontal collaborative capabilities (symmetric stations execute the same working mode) and vertical collaborative capabilities (dynamic parameter adjustment of the entire side of the robots) through Internet of Things chips; the welding working mode defined by the welding current and wire feeding speed is pre-configured, and a wooden barrel dynamic model with the longitudinal chain as the bottom surface and the horizontal branch chain as the top surface is constructed to represent the collaborative network; the dynamic balance value is generated by calculating the horizontal entropy difference and the vertical entropy flow, and the welding parameters are optimized under the process constraints of penetration and weld width;

[0062] The present invention significantly improves the collaborative efficiency and welding quality stability of welding robots and reduces the process error through the Internet of Things and group intelligent technology, as reflected in:

[0063] Improved collaborative efficiency: Through the three-dimensional collaborative network constructed by the vertical and horizontal Internet of Things chains, horizontal synchronization (reducing parameter errors in symmetric workstations) and vertical compensation (reducing parameter transfer delays between workstations) of welding robots are achieved. Compared with traditional single robots, the control efficiency is improved, especially suitable for the scenario of segmented welding of long welds;

[0064] Enhanced process adaptability: The pre-configured working mode is combined with dynamic constraints on penetration and bead width, supporting automatic parameter switching for different materials such as stainless steel (penetration coefficient 0.08mm / A) and aluminum alloy (penetration coefficient 0.05mm / A), etc., without manual intervention, which is conducive to shortening production changeover time;

[0065] Optimized quality stability: Based on the calculation of collaborative entropy difference and feedback optimization of the barrel model, the weld defect rate (such as lack of fusion, crack) is reduced, and the fluctuations of penetration and bead width are controlled within the tolerance range (such as penetration tolerance ±0.2mm, bead width tolerance ±1.0mm), meeting the high-precision welding requirements such as in aerospace;

[0066] Strong system scalability: The Internet of Things architecture supports modular expansion. Newly added welding robots can be quickly connected to the existing collaborative network through the vertical chain and horizontal branch chain, adapting to the dynamic adjustment of the production line scale and reducing the system upgrade cost. Description of the Drawings

[0067] The drawings are used to provide a further understanding of the present invention, and constitute a part of the specification. They are used together with the embodiments of the present invention to explain the present invention, and do not constitute a limitation to the present invention.

[0068] Figure 1 It is a step schematic diagram of a method for group intelligent control of welding robots based on the Internet of Things according to the present invention. Detailed Embodiment

[0069] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative work shall fall within the protection scope of the present invention.

[0070] In the first embodiment: A group intelligent control system for welding robots based on the Internet of Things is provided. The system includes: a workstation layout and network construction module, a collaborative ability assignment module, a working mode configuration and model construction module, and a collaborative control and parameter optimization module.

[0071] The workstation layout and network construction module sets the workstation sequence on both sides of the assembly line based on the welding process logic, forms a three-dimensional parallel operation group with horizontally dislocated and symmetrical welding robots, and establishes a vertical feedback IoT chain and a horizontal feedback IoT branch chain;

[0072] Exemplarily, the workstation layout and network construction module includes a job parallel group construction unit and a feedback network setting unit;

[0073] The parallel operation group construction unit is used to form a parallel operation group with two welding robots that are horizontally displaced and symmetrical on both sides of the production line to perform segmented welding of the same welding object area;

[0074] The feedback network setting unit is used to set up a longitudinal feedback Internet of Things chain between the welding robots on one side of the assembly line based on the workstation sequence, and to set up a transverse feedback Internet of Things branch chain between the symmetrical robots in the parallel operation group.

[0075] The collaborative capability granting module grants the welding robot horizontal and vertical collaborative capabilities through horizontal feedback IoT branches and vertical feedback IoT chains;

[0076] Exemplarily, the collaborative capability granting module includes a horizontal collaborative control unit and a vertical collaborative control unit;

[0077] The horizontal collaborative control unit enables the symmetrical welding robots in the parallel operation group to perform the same welding work mode through the horizontal feedback of the IoT branch chain and the IoT chip;

[0078] The longitudinal collaborative control unit dynamically adjusts the welding working mode of all welding robots on one side through the longitudinal feedback IoT chain and IoT chip.

[0079] The working mode configuration and model building module is used to preconfigure the welding working mode and build a barrel dynamic model to characterize the robot collaborative network;

[0080] Exemplarily, the working mode configuration and model building module includes a welding working mode preconfiguration unit and a barrel dynamic model building unit;

[0081] A welding working mode pre-configuration unit, used to define a welding working mode consisting of a welding current and a wire feeding speed;

[0082] The barrel dynamic model construction unit is used to use the longitudinal feedback IoT chain as the bottom surface of the barrel dynamic model and the transverse feedback IoT branch chain as the top surface of the barrel dynamic model, and to connect the bottom surface and the top surface through the symmetry of the parallel operation group to form a barrel dynamic model that represents the robot collaborative network.

[0083] A collaborative control and parameter optimization module is used to establish a swarm intelligence control collaborative model and a swarm intelligence control feedback model, and optimize the welding current and wire feeding speed in real time to meet the process constraints of penetration and bead width.

[0084] Exemplarily, the collaborative control and parameter optimization module includes a collaborative model calculation unit and a feedback optimization unit.

[0085] The collaborative model calculation unit is used to record the real-time operation state vectors of the welding robots at each station, and calculate the horizontal entropy difference and the vertical entropy flow.

[0086] The feedback optimization unit is used to map the horizontal entropy difference and the vertical entropy flow to a wooden barrel dynamic model, generate a dynamic balance value, and minimize the balance value under the constraints of penetration tolerance and bead width tolerance, where the penetration tolerance and bead width tolerance are determined according to the maximum and minimum values of penetration and bead width required by the welding process.

[0087] Please refer to Figure 1 , in the second embodiment: A swarm intelligence control method for a welding robot group based on the Internet of Things is provided to be applicable to the first embodiment above. The method includes the following steps:

[0088] Step S1: Set the station order on both sides of the assembly line based on the welding process logic, form a three-dimensional operation parallel group with two horizontally misaligned and symmetric welding robots, and establish a dynamic communication connection between the robots through a longitudinal feedback Internet of Things chain and a horizontal feedback Internet of Things branch chain.

[0089] Exemplarily, step S11: Set the station order according to the welding process logic. The stations have horizontal misalignment symmetry on both sides of the assembly line. Form an operation parallel group with two horizontally misaligned and symmetric welding robots, and perform segmented welding on the same welding object area at the stations.

[0090] Step S12: Select all the welding robots on one side of the assembly line, set the longitudinal feedback Internet of Things chain of the welding robots based on the station order, and set the horizontal feedback Internet of Things branch chain between the two horizontally misaligned and symmetric welding robots.

[0091] Step S2: Based on the horizontal feedback Internet of Things branch chain, endow the two horizontally misaligned and symmetric welding robots with horizontal collaborative capabilities so that they execute the same welding working mode; based on the longitudinal feedback Internet of Things chain, endow the welding robots on the same side with longitudinal collaborative capabilities to dynamically adjust the welding parameters of the entire side of the robots.

[0092] Exemplarily, step S21: Based on the horizontal feedback IoT branch chain, through the IoT chip, two welding robots with a horizontal feedback IoT branch chain relationship are made to have horizontal collaboration capabilities, and the horizontal collaboration capabilities are used to enable two welding robots with horizontal misalignment symmetry to execute the same welding work mode;

[0093] Step S22: Based on the vertical feedback IoT chain, through the IoT chip, one side of the welding robots with a work station sequence relationship is made to have vertical collaboration capabilities, and the vertical collaboration capabilities are used to adjust the welding work modes of all the welding robots on one side of both sides of the assembly line.

[0094] Step S3: Pre-configure multiple groups of welding work modes, where the modes are defined by welding current and wire feeding speed, and construct a wooden barrel dynamic model to represent the robot collaboration network. Among them, the bottom surface is the vertical feedback IoT chain, the top surface is the horizontal feedback IoT branch chain, and the bottom surface and the top surface are connected symmetrically through the job parallel group;

[0095] Exemplarily, step S31: Preset several welding work modes differentiated according to welding process requirements. The welding work modes are configured by welding current and wire feeding speed, and any i-th welding work mode is denoted as M i ={I i ,SV i}, where I i and SV i respectively represent the welding current and wire feeding speed in the i-th welding work mode;

[0096] Step S32: Construct a wooden barrel dynamic model for representing the dynamic collaboration network among welding robots. The bottom surface of the wooden barrel dynamic model is the vertical feedback IoT chain, the top surface of the wooden barrel dynamic model is the horizontal feedback IoT branch chain, and the connection relationship between the bottom surface and the top surface is the symmetry relationship of the job parallel group based on the work stations on both sides of the assembly line. Then, for one side of the welding robots U x at the x-th work station, there is a connection edge F x ={U x →U x+1 ,U x →U x‘} in the wooden barrel dynamic model. Among them, U x →U x+1 exists in the vertical feedback IoT chain, U x →U x‘ exists in the horizontal feedback IoT branch chain, and U x‘ is the other side of the welding robot at the x-th work station in the job parallel group.

[0097] Step S4: Establish a swarm intelligence control collaboration model, collect the operating state parameters of each welding robot in real time, calculate the horizontal entropy difference and vertical entropy flow, map them to the wooden barrel dynamic model to generate a dynamic balance value, and minimize this balance value by optimizing the welding current and wire feeding speed under the process constraint conditions of penetration and bead width;

[0098] Exemplarily, step S41: Construct a swarm intelligence control collaboration model, including a horizontal collaboration model and a vertical collaboration model:

[0099] Based on the working mode of the welding machine, record the operating state vector of the welding robot U on one side of the xth station x of The operating state vector of the welding robot U on the other side x‘ of t is the independent variable of the time node, and are respectively the real-time welding current and wire feeding speed of the welding robot U operating in the ith welding working mode at the time node independent variable t x ; and are respectively the real-time welding current and wire feeding speed of the welding robot U operating in the ith welding working mode at the time node independent variable t x‘ ;

[0100] Horizontal collaboration model:

[0101]

[0102] Vertical collaboration model:

[0103]

[0104] In the formula, H h (x) represents the horizontal entropy difference, and H z (x) represents the vertical entropy flow. t 1 is the end time within a sampling period, and t 0 is the start time within a sampling period;

[0105] Step S42: Based on the swarm intelligence control collaboration model, construct a swarm intelligence control feedback model. Take the horizontal entropy difference H h (x) as the side length of the connection edge U x →U x‘ , and take the vertical entropy flow H z (x) as the side length of the connection edge U x →U x+1The side lengths are mapped into the dynamic model of the wooden barrel to generate the volume of the dynamic model of the wooden barrel, which is used as the dynamic balance value of the wooden barrel. The volume function is denoted as E;

[0106] Table 1 Simulation results of partial parameters:

[0107]

[0108]

[0109] Adjust the welding current and wire feeding speed through the welding process constraint conditions to minimize the dynamic balance value of the wooden barrel:

[0110]

[0111] In the formula, D(x, t) is the empirical penetration linear function, and K D is the penetration coefficient, W(x, t) is the empirical bead width linear function, and K W is the bead width coefficient, b D and b W are both overshoot control coefficients; D 0 is the target penetration, and W 0 is the target bead width, and D max and D min are the maximum and minimum penetration values required by the welding process respectively, and W max and W min are the maximum and minimum bead width values required by the welding process respectively; δ D is the penetration tolerance, and δ W is the bead width tolerance, and

[0112] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or elements inherent to such process, method, article or device.

[0113] Finally, it should be noted that the above are only the preferred embodiments of the present invention and are not used to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions recorded in the foregoing embodiments or perform equivalent replacements on some of the technical features. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A welding robot swarm intelligent control method based on the Internet of Things, characterized in that: The method comprises the following steps: Step S1: The workstation sequence on both sides of the assembly line is set based on the welding process logic, and two horizontally dislocated and symmetrical welding robots are formed into a three-dimensional parallel operation group, and a dynamic communication connection between the robots is established through the longitudinal feedback IoT chain and the transverse feedback IoT branch chain; Step S2: Based on the lateral feedback IoT branch chain, two horizontally dislocated and symmetrical welding robots are given lateral coordination capability so that they can perform the same welding working mode; based on the longitudinal feedback IoT chain, the welding robots on the same side are given longitudinal coordination capability so as to dynamically adjust the welding parameters of the robots on the entire side; Step S3: pre-configure multiple groups of welding working modes, where the modes are defined by welding current and wire feeding speed, and construct a barrel dynamic model to represent the robot collaborative network, wherein the bottom surface is a longitudinal feedback IoT chain, the top surface is a transverse feedback IoT branch chain, and the bottom surface and the top surface are connected by the symmetry of the parallel operation group; Step S4: Establish a swarm intelligent control collaborative model, collect the operating status parameters of each welding robot in real time, calculate the transverse entropy difference and longitudinal entropy flow, map them to the barrel dynamic model to generate a dynamic balance value, and minimize the balance value by optimizing the welding current and wire feeding speed while satisfying the process constraints of the depth and width of penetration.

2. The method for intelligent control of welding robots based on the Internet of Things according to claim 1 is characterized in that: The specific implementation process of step S1 includes: Step S11: setting the workstation sequence according to the welding process logic, wherein the workstations have horizontal misalignment symmetry on both sides of the assembly line, forming a parallel operation group with two welding robots having horizontal misalignment symmetry, and performing segmented welding of the same welding object area at the workstations; Step S12: Select all welding robots on one side of the two sides of the assembly line, and set up a longitudinal feedback Internet of Things chain of the welding robots based on the workstation sequence, and set up a transverse feedback Internet of Things branch chain between the two welding robots based on two welding robots with horizontal misalignment symmetry.

3. The method for intelligent control of welding robots based on the Internet of Things according to claim 1 is characterized in that: The specific implementation process of step S2 includes: Step S21: Based on the horizontal feedback IoT branch, two welding robots in the horizontal feedback IoT branch relationship are enabled to have horizontal coordination capability through the IoT chip, wherein the horizontal coordination capability is used to enable two welding robots with horizontal misalignment symmetry to perform the same welding working mode; Step S22: Based on the longitudinal feedback IoT chain, the IoT chip enables the welding robots on one side with a workstation sequence relationship to have longitudinal coordination capability, and the longitudinal coordination capability is used to adjust the welding working mode of all welding robots on one side of the two sides of the assembly line.

4. The method for intelligent control of welding robots based on the Internet of Things according to claim 1 is characterized in that: The specific implementation process of step S3 includes: Step S31: presetting a number of welding working modes differentiated according to welding process requirements, wherein the welding working modes are configured by welding current and wire feeding speed, and any i-th welding working mode is recorded as M i = {I i , SV i }, where I i and SV i They represent the welding current and wire feeding speed in the i-th welding working mode respectively; Step S32: Construct a barrel dynamic model to characterize the dynamic collaborative network between welding robots. The bottom surface of the barrel dynamic model is a longitudinal feedback IoT chain, and the top surface of the barrel dynamic model is a transverse feedback IoT branch chain. The connection relationship between the bottom surface and the top surface is based on the symmetry relationship of the parallel operation groups of the workstations on both sides of the assembly line. For the welding robot U on one side of the x-th workstation, x , in the barrel dynamic model there is a connecting edge F x = {U x →U x+1 , U x →U x‘ }, where U x →U x+1 Existing in the vertical feedback IoT chain, U x →U x‘ Existing in the horizontal feedback IoT branch, U x‘ It is the welding robot on the other side of the xth workstation in the parallel operation group.

5. The method for intelligent control of welding robots based on the Internet of Things according to claim 4 is characterized in that: The specific implementation process of step S4 includes: Step S41: construct a swarm intelligent control collaborative model, including a horizontal collaborative model and a vertical collaborative model: Based on the welding machine working mode, record the welding robot U on one side of the x-th station x The running state vector The other side welding robot U x‘ The running state vector t is the time node independent variable, and are respectively the welding robot U running in the i-th welding working mode at the time node independent variable t x Real-time welding current and wire feed speed, and are respectively the welding robot U running in the i-th welding working mode at the time node independent variable t x‘ Real-time welding current and wire feed speed; Horizontal collaboration model: Vertical collaboration model: In the formula, H h (x) represents the transverse entropy difference, H z (x) represents the longitudinal entropy flow, t1 is the end time of a sampling period, t0 is the start time of a sampling period, Step S42: Based on the swarm intelligent control collaborative model, a swarm intelligent control feedback model is constructed, and the lateral entropy difference H h (x) as the connecting edge U x →U x‘ The length of the side will be the longitudinal entropy flow H z (x) as the connecting edge U x →U x+1 The side length is mapped to the barrel dynamic model to generate the volume of the barrel dynamic model as the barrel dynamic balance value. The volume function is recorded as E. Through the constraints of welding process, the dynamic balance value of the barrel is minimized by adjusting the welding current and wire feeding speed: Where D(x, t) is the empirical linear function of penetration depth, and K D is the penetration coefficient, W(x, t) is the empirical linear function of the weld width, and K W is the melt width coefficient, b D and b W are all overshoot control coefficients; D0 is the target melting depth, and W0 is the target weld width, and D max and D min are the maximum and minimum penetration depths required by the welding process, W max and W min are the maximum and minimum weld widths required by the welding process; δ D is the penetration tolerance, and δ W is the weld width tolerance, and 6. A welding robot swarm intelligent control system based on the Internet of Things, executing the welding robot swarm intelligent control method as claimed in claim 1, characterized in that: The system includes: a workstation layout and network construction module, a collaborative capability granting module, a work mode configuration and model construction module, and a collaborative control and parameter optimization module; The station layout and network construction module sets the station sequence on both sides of the assembly line based on the welding process logic, forms a three-dimensional parallel operation group with horizontally dislocated and symmetrical welding robots, and establishes a vertical feedback IoT chain and a horizontal feedback IoT branch chain; The collaborative capability granting module grants the welding robot horizontal collaborative capability and vertical collaborative capability through the horizontal feedback IoT branch chain and the vertical feedback IoT chain; The working mode configuration and model building module is used to preconfigure the welding working mode and build a barrel dynamic model to characterize the robot collaborative network; The collaborative control and parameter optimization module is used to establish a swarm intelligent control collaborative model and a swarm intelligent control feedback model, and optimize the welding current and wire feeding speed in real time to meet the process constraints of the depth of penetration and the width of penetration.

7. The welding robot swarm intelligent control system based on the Internet of Things according to claim 6 is characterized in that: The workstation layout and network construction module includes a job parallel group construction unit and a feedback network setting unit; The parallel operation group construction unit is used to form a parallel operation group with two welding robots that are horizontally displaced and symmetrical on both sides of the assembly line to perform segmented welding of the same welding object area; The feedback network setting unit is used to set up a longitudinal feedback Internet of Things chain between welding robots on one side of the assembly line based on the workstation sequence, and to set up a transverse feedback Internet of Things branch chain between symmetrical robots in the parallel operation group.

8. The welding robot swarm intelligent control system based on the Internet of Things according to claim 6 is characterized in that: The cooperation capability granting module includes a horizontal cooperation control unit and a vertical cooperation control unit; The horizontal collaborative control unit enables the symmetrical welding robots in the parallel operation group to perform the same welding working mode through horizontal feedback of the IoT branch chain and the IoT chip; The longitudinal collaborative control unit dynamically adjusts the welding working mode of all welding robots on a single side through the longitudinal feedback Internet of Things chain and the Internet of Things chip.

9. The welding robot swarm intelligent control system based on the Internet of Things according to claim 6 is characterized in that: The working mode configuration and model building module includes a welding working mode pre-configuration unit and a barrel dynamic model building unit; The welding working mode pre-configuration unit is used to define the welding working mode composed of welding current and wire feeding speed; The barrel dynamic model construction unit is used to use the longitudinal feedback Internet of Things chain as the bottom surface of the barrel dynamic model and the transverse feedback Internet of Things branch chain as the top surface of the barrel dynamic model, and symmetrically connect the bottom surface and the top surface through the parallel operation group to form a barrel dynamic model that represents the robot collaborative network.

10. The welding robot swarm intelligent control system based on the Internet of Things according to claim 6, characterized in that: The collaborative control and parameter optimization module includes a collaborative model calculation unit and a feedback optimization unit; The collaborative model calculation unit is used to record the real-time operation state vector of the welding robot at each station and calculate the transverse entropy difference and the longitudinal entropy flow; The feedback optimization unit is used to map the transverse entropy difference and the longitudinal entropy flow to the barrel dynamic model, generate a dynamic balance value, and minimize the balance value under the constraints of the penetration tolerance and the weld width tolerance, wherein the penetration tolerance and the weld width tolerance are determined according to the maximum and minimum values ​​of the penetration depth and the weld width required by the welding process.

Images