A laser tracking based welding control method and system

Abstract

The application provides a welding control method and system based on laser tracking, and the method comprises the following steps: extracting a welding seam deviation result collected by a laser sensor; performing reliability determination on the welding seam deviation result to form a deviation credibility result; generating a candidate control instruction based on the deviation credibility result; and performing execution constraint on the candidate control instruction and delivering the candidate control instruction to an execution mechanism. By introducing a deviation credibility determination mechanism, the application effectively identifies and suppresses the false operation caused by sensor interference, and solves the technical problem of poor welding process stability in the prior art. The scheme can ensure welding precision while significantly improving the robustness and safety of the welding process.

Description

Technical Field

[0001] This invention belongs to the field of industrial control, and in particular relates to a welding control method and system based on laser tracking. Background Technology

[0002] In automated welding equipment, laser-tracking-based weld seam recognition and control technology has become an important means to improve welding consistency, reduce manual intervention, and adapt to assembly errors in complex workpieces. Existing solutions typically involve placing a line laser sensor, laser displacement sensor, or structured light camera in front of the welding torch, allowing the laser line to span the area to be welded. This enables real-time acquisition of information on the weld seam profile, bevel edge, weld center, or relative deviation. The control system then corrects the welding torch's lateral position, trajectory, or posture based on the detection results, thereby achieving weld seam tracking control. This approach has a clear engineering foundation and can achieve good automatic tracking in scenarios with regular straight weld seams, small assembly errors, stable workpiece surface conditions, and weak welding interference.

[0003] However, in actual engineering welding processes, especially in arc welding, thick plate bevel welding, complex curved welds, misaligned assemblies, fluctuating bevel gaps, highly reflective metal surfaces, and conditions with significant spatter, the weld profile obtained by laser tracking is often not a stable, complete, and continuous ideal profile. Strong arc light, spatter particles, fume obstruction, molten pool reflection, bevel edge burrs, and localized workpiece contamination can all cause localized peaks, missing boundaries, center drift, or short-term jumps in the laser profile. Existing control methods typically use the obtained weld center or deviation as the basis for welding torch correction, rarely distinguishing whether the deviation originates from the actual bevel structure, and rarely combining short-term continuous deviation changes to determine whether they represent the actual weld trend. In this way, when a single frame profile is affected by spatter or reflection, the controller can easily identify abnormal deviations as actual weld seam offsets and generate large lateral correction actions accordingly. When candidate control actions are further directly input into the actuator, excessive changes in single-cycle actions may cause lateral oscillation of the welding torch, offset of the arc point of action, and fluctuation of the molten pool edge, ultimately leading to problems such as weld misalignment, uneven weld bead formation, weld beads, undercut, or local lack of fusion. Welding control differs from general position servo control. Changes in the lateral position of the welding torch simultaneously affect the arc point of action, molten pool stability, and weld bead formation. Therefore, once detection errors are transmitted to the actuator, they can easily be amplified into actual quality defects during continuous welding. While existing technologies possess basic capabilities such as laser detection, deviation extraction, and welding torch correction, they still lack an improved mechanism for step-by-step constraints along the laser tracking control link under complex welding conditions. Specifically, they lack the ability to determine whether weld deviations originate from effective groove structures, to reliably quantify the short-term continuity and abrupt changes of deviations, and to constrain the continuity of action before candidate control commands enter the actuator. Therefore, it is difficult to suppress the problems of miscorrection and overcorrection caused by contour distortion from the three stages of perception, decision-making, and execution. Summary of the Invention

[0004] This invention discloses a welding control method and system based on laser tracking to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, a first aspect of the present invention provides a welding control method based on laser tracking, the method comprising: Acquire weld contour data and extract weld deviation results; The reliability of the weld deviation results is determined to form a deviation confidence result; the steps of determining the reliability of the weld deviation results include: A short-time deviation sequence is introduced to compare the weld deviation of the current frame with its historical changes; The variation characteristics of the deviation sequence are described based on the maximum deviation amplitude and discrete second-order difference, and the deviation confidence result is calculated by combining the weld structure identification. The steps for calculating the deviation confidence result based on the maximum deviation amplitude and discrete second-order difference to describe the variation characteristics of the deviation sequence, and combined with weld structure identification, include: Calculate the maximum deviation magnitude of the weld deviation in the current frame relative to the weld deviation in historical frames; Calculate the discrete second-order difference between the weld deviation in the current frame and the weld deviation in the historical frames; The maximum deviation amplitude and discrete second-order difference are fused with the weld structure identifier to obtain the deviation confidence result, wherein the weld structure identifier is determined based on the groove boundary spacing and the continuity of the contour change. Candidate control instructions are generated based on the deviation confidence results; The candidate control instructions are subject to execution constraints, and the execution control instructions are issued.

[0006] Furthermore, the step of acquiring weld contour data and extracting weld deviation results includes: The weld contour data is obtained by the laser tracking module. The weld search window is captured with the reference center position of the welding gun. The bevel structure is found in the window, and then the left bevel boundary and the right bevel boundary are identified. The midpoint between the left and right bevel boundaries is calculated as the geometric center of the weld. The geometric center of the weld is then compared with the reference center of the welding torch to obtain the weld deviation result.

[0007] Furthermore, the step of identifying the left bevel boundary and the right bevel boundary includes: The position where the profile transitions from the flat area of ​​the parent material into the continuously descending area is defined as the left bevel boundary. The right bevel boundary is defined as the point where the contour rises continuously from the bevel area and then flattens out again. If multiple sets of candidate left and right boundaries that meet the conditions are formed in the current search window, the set whose midpoint is closest to the geometric center of the weld confirmed in the previous control cycle shall be selected as the left bevel boundary and the right bevel boundary.

[0008] Furthermore, the step of generating candidate control instructions based on the deviation confidence result includes: The deviation confidence result is used as the confidence participation weight of the current deviation; The deviation confidence result is combined with the weld deviation result of the current frame to generate candidate control commands, so that the welding torch maintains smooth operation when the deviation confidence is low or the target changes too much.

[0009] Furthermore, the step of using the deviation confidence result as the confidence participation weight of the current deviation includes: The target control quantity is calculated by combining the deviation confidence result with the weld deviation result of the current frame. The difference between the target control quantity and the actual control command executed in the previous cycle is calculated as the target difference. The normalized magnitude of the target difference is placed in the denominator. When the target difference is large, the denominator value is increased to compress the control increment.

[0010] Furthermore, the step of imposing execution constraints on the candidate control instructions and issuing execution control instructions includes: The candidate control instruction is compared with the execution control instruction actually issued in the previous cycle; When the change range of the candidate control instruction exceeds the allowable change range in a single cycle, the candidate control instruction is truncated as the sum of the execution control instruction actually issued in the previous cycle and the allowable change range in a single cycle, and an execution control instruction is formed.

[0011] Furthermore, after the step of imposing execution constraints on the candidate control instructions and issuing execution control instructions, the method further includes: The execution control command after the welding torch actuator completes its action is recorded as the previous execution control command for the next control cycle; If no position feedback or servo response is received within the preset feedback time, the previously confirmed executed control command is retained as the starting execution value for the next control cycle.

[0012] In a second aspect, the invention provides a laser-tracking-based welding control system, the system comprising: Laser contour acquisition unit, configured to acquire weld contour data; A bevel structure recognition unit is configured to extract weld deviation results. A deviation reliability determination unit is configured to determine the reliability of the weld deviation result and generate a deviation reliability result; the step of determining the reliability of the weld deviation result includes: A short-time deviation sequence is introduced to compare the weld deviation of the current frame with its historical changes; The variation characteristics of the deviation sequence are described based on the maximum deviation amplitude and discrete second-order difference, and the deviation confidence result is calculated by combining the weld structure identification. The steps for calculating the deviation confidence result based on the maximum deviation amplitude and discrete second-order difference to describe the variation characteristics of the deviation sequence, and combined with weld structure identification, include: Calculate the maximum deviation magnitude of the weld deviation in the current frame relative to the weld deviation in historical frames; Calculate the discrete second-order difference between the weld deviation in the current frame and the weld deviation in the historical frames; The maximum deviation amplitude and discrete second-order difference are fused with the weld structure identifier to obtain the deviation confidence result, wherein the weld structure identifier is determined based on the groove boundary spacing and the continuity of the contour change. A candidate control instruction generation unit is configured to generate candidate control instructions based on the deviation confidence result.

[0013] The beneficial technical effects of the present invention are at least as follows: To address the aforementioned problems, this invention provides a welding control method and system based on laser tracking. The core of this method involves first constraining the weld contour obtained through laser tracking to form weld deviation results and weld structure identifiers using bevel structure constraints. Then, combining this with a short-time deviation sequence, the continuity, abrupt change, and structural effectiveness of the current deviation are assessed for reliability, resulting in a deviation confidence score that can participate in control weight allocation. Subsequently, the control system uses the weld deviation results and the deviation confidence score in pairs. When generating candidate control commands, it simultaneously considers the weld offset direction, offset amplitude, the reliability of the deviation information, and the actual control quantity executed in the previous cycle. This allows the welding torch correction action to gradually converge with the actual weld offset and automatically reduces the response intensity when low-confidence or abrupt deviations occur. Finally, the candidate control commands are constrained according to the actuator's motion capability and the continuity of actions allowed by the welding process, forming actual execution control commands issued to the welding torch actuator. This enables the welding torch to complete a restricted, smooth, and executable lateral correction within each control cycle, and the final execution control command and execution status for this cycle are written into the control record. This invention can improve the stability of weld tracking under complex working conditions and the controllability of welding torch correction actions while retaining the hardware foundation of existing laser tracking welding systems, and reduce the risks of weld deviation, weld beads and uneven weld formation caused by abnormal contours, pseudo deviations and control jumps. Attached Figure Description

[0014] The present invention will be further described with reference to the accompanying drawings, but the embodiments in the drawings do not constitute any limitation on the present invention. For those skilled in the art, other drawings can be obtained based on the following drawings without creative effort.

[0015] Figure 1 This is a flowchart of a laser tracking-based welding control method according to the present invention.

[0016] Figure 2 This is a framework diagram of a laser-tracking-based welding control system according to the present invention. Detailed Implementation

[0017] Embodiments of the present invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0018] In one or more embodiments, such as Figure 1 As shown, a welding control method based on laser tracking is disclosed, the method comprising the following: S1: Obtain weld contour data and extract weld deviation results.

[0019] Specifically, step one involves the laser tracking module acquiring weld contour data and generating weld deviation results and weld structure identifiers for subsequent steps. The laser tracking module is mounted in front of the welding torch, allowing the line laser to traverse the bevel area to be welded; each sample from the sensor yields a transverse contour sequence, denoted as... .in, The sampling position is along the laser scanning direction. This represents the contour height value corresponding to the sampling location. This data is output by a line laser sensor or structured light camera through an industrial control interface and is presented as a one-dimensional array arranged in the scanning order. The welding torch reference center position is denoted as... The current trajectory control position is given by the welding robot or welding torch actuator and mapped to the laser contour coordinates through calibration before welding. The system uses... The weld search window is centered, and the bevel structure is located within this window. Then, the left bevel boundary is identified. and the right bevel boundary The left bevel boundary is determined by the position where the profile transitions from the flat area of ​​the base material to the continuously descending area, while the right bevel boundary is determined by the position where the profile continuously rises from the bevel area and returns to a flat state. In practice, the height changes of adjacent sampling points are compared. If several consecutive sampling points show a slope change in the same direction, and the change section corresponds to the flat area of ​​the base material and the bevel area respectively, then this position is confirmed as a candidate boundary. If a local sampling point shows a peak but the preceding and following contours lack a continuous bevel shape, it is removed during boundary confirmation. If multiple sets of candidate left and right boundaries that meet the conditions are formed within the current search window, the set whose midpoint is closest to the confirmed weld geometric center of the previous control cycle is selected first. The confirmed weld geometric center of the previous control cycle corresponds to the midpoint of the previous valid left and right bevel boundaries. If there is no previously confirmed weld geometric center, the set whose midpoint is closest to the previous one is selected. A set; if the current frame fails to simultaneously identify valid left and right bevel boundaries, or if the calibration mapping is abnormal, resulting in... , , If they cannot be in the same coordinate system, the frame is recorded as an invalid contour frame. This ensures that the boundary position corresponds more to the geometry of the bevel itself, rather than to local spatter or abrupt changes in reflection during the welding process.

[0020] The calculation of weld deviation results is derived from the fundamental relationship of "midpoint coordinates of two points" in analytical geometry. For a laser profile spanning the bevel, when the left and right bevel boundaries are respectively... and At this point, the geometric center of the weld is represented by the midpoint of the two boundaries; then this center is compared with the welding torch reference center. The comparison yields the deviation of the weld relative to the welding torch. Based on this midpoint relationship, this application limits the two points used for calculation to those selected after continuous bevel structure screening. and This makes the deviation results more closely match the actual bevel boundary. The weld deviation results are obtained by the following formula: ; in, The weld deviation result is calculated in this step. This is the location of the left bevel boundary, by The location where the middle section transitions from a flat zone to a continuously descending zone is determined; This is the location of the right-side bevel boundary, by The location is determined by the continuous rise from the slope area and the subsequent return to a gentle slope. The reference center position of the welding torch is obtained through calibration from the current trajectory position of the welding actuator. In this formula, the geometric center of the weld is first determined by the left and right bevel boundaries, and then the deviation is obtained from the difference between the geometric center of the weld and the reference center of the welding torch; if the system uses normalized sampling coordinates, then... , , First, convert all frames to the same normalized coordinates before substituting them into the calculation. If the current frame is recorded as an invalid contour frame, then the current period's... Use the most recent valid weld deviation result; if there are no valid historical deviation results at the initial stage of system startup, then use the current cycle. according to The process is handled, and an invalid profile flag is written to the controller. Taking a single frame of laser profile as an example, the left bevel boundary identified within the search window... Right side bevel boundary Welding torch reference center position The center of the weld is ,therefore This indicates that the weld center is aligned with the welding torch reference center; when identified in another frame... , , The center of the weld is , This indicates that the weld center has shifted relative to the welding torch reference center.

[0021] In obtaining Subsequently, the system synchronously generates weld structure identifiers. , is used to indicate whether the current profile has a basic weld structure. It is determined by three engineering conditions: First, and The distance between them is within the preset bevel range; secondly, and The contour changes in the vicinity are continuous; thirdly, there is a concave or bevel transition shape between the left and right boundaries relative to the parent material on both sides. If all three conditions are met, then ;otherwise For the aforementioned invalid contour frames, directly set... For example, when the distance between two boundaries is detected to be significantly smaller than the preset process range, and there is no recessed structure in between, even if the distance can be calculated based on the boundary positions... , will also Set as When the distance between the left and right boundaries falls within the preset process range, and the middle area exhibits a continuous bevel shape, then... Set as This step ultimately outputs the weld deviation result. and weld structure markings The next step is to receive... and This allows us to determine the reliability of the current weld deviation results.

[0022] S2: Determine the reliability of the weld deviation results to form a deviation credibility result.

[0023] Specifically, this step uses the weld deviation results output in step one. and weld structure markings Using this as the basis, the reliability of the current deviation results is determined, and a deviation credibility result is generated. During welding, the spatial positional changes of the actual weld are usually continuous, and its deviation sequence exhibits a smooth change over a short period of time. However, interference caused by spatter, reflection, or local obstruction manifests as sudden jumps in the deviation or abnormal trends in its change. Therefore, this step, based on the geometric deviation obtained in step one, introduces short-time series analysis to compare the current deviation with its historical changes, thereby determining whether the deviation conforms to the continuous geometric characteristics of the weld.

[0024] During implementation, the controller stores the output from step one within a continuous sampling period. The short-time deviation sequence is constructed, and the deviation of the current frame is denoted as... Historical frame deviation is denoted as , The sequence length is denoted as... These data all originate from continuous sampling results from the laser tracking module and maintain the same spatial coordinate system as in step one. The preset sequence length has not yet been accumulated during the startup phase. When the current frame only has valid offset results, a short-time offset sequence is constructed using all currently cached valid offset results; if the current frame only has valid offset results, the maximum offset term is calculated as follows: If there are fewer than two valid historical frames, the second-order difference term is processed according to... The data is processed, and available historical frames are used in the remaining calculations. Based on discrete sequence analysis methods in mathematics, the "maximum deviation magnitude" and "discrete second-order difference" are used to jointly describe the changing characteristics of the deviation sequence. The maximum deviation magnitude, derived from range analysis, describes the maximum change in the current deviation relative to historical deviations; the discrete second-order difference, derived from numerical analysis of sequence curvature, describes the degree of abrupt change in the deviation trend. These two parts are then integrated with the structural identifier. The results are then fused to obtain the bias confidence level.

[0025] Deviation Reliability The calculation form is as follows: ; in, This is the result of the bias confidence level; This is a weld structure identifier, indicating whether the current profile meets the weld groove structure requirements; This refers to the weld deviation in the current frame. For historical frame weld deviations; The length of the short time series; This serves as a normalized baseline value for deviation variation, determined by the weld groove width or allowable deviation range. It is used to compress deviation variations to a uniform scale. Take greater than If the value is positive and the corresponding parameter has not been updated after the current process switch, then the value corresponding to the previously confirmed process will be used. ; This is the trend weighting coefficient, used to adjust the impact of trend abrupt changes on credibility. The first term "1" in the denominator of the formula comes from the basic form of normalized weights, used to avoid the denominator being zero; the second term... Derived from range normalization, it describes the maximum percentage change in the current deviation relative to historical deviations; the third term Derived from discrete second-order difference, its physical meaning corresponds to the "curvature" of the deviation change; a large value indicates a significant abrupt change in the deviation. This expression is based on classical range analysis and the second-order difference method, and introduces structural identifiers. and normalization coefficient Improve the integration process to make it suitable for determining the contour data that is disturbed during the welding process.

[0026] In the embodiments of this application, the deviation confidence result The value range of [[ ]] is from 0 to 1, and its numerical value directly reflects the reliability and change state of the weld deviation data of the current frame. The specific determination logic is as follows: When C = 0, it means that the current contour does not meet the requirements of the weld structure, or the current deviation does not have the reliability as a control basis; this indicates that the detected deviation is very likely to be an abnormal value caused by arc light interference, splash occlusion or sensor noise, and does not have the basis of the weld structure. At this time, the control system will directly eliminate the deviation data of this frame and maintain the control state of the previous moment to avoid misoperation.

[0027] When 0 < C < 0.25, it means that there is an obvious jump or trend mutation in the current deviation relative to the short-term deviation sequence, usually corresponding to a low-confidence deviation caused by splash, reflection, occlusion or local abnormality; When 0.25 ≤ C < 0.5, it means that the current deviation has a certain continuity, but there are still obvious fluctuations. It is advisable to reduce its participation degree during subsequent control; although the data is not completely distorted at this time, the stability decreases. The control system can introduce a damping coefficient or perform smoothing filtering on the deviation result to prevent the welding torch from shaking; When C ≥ 0.5, it means that the current deviation maintains a good continuous relationship with the short-term deviation sequence and can be used as the main basis for correcting the welding torch; at this time, the deviation data is within the stable change range and no abnormal mutation occurs. The control system will generate a normal welding deviation correction instruction based on this deviation result.

[0028] In the specific calculation process, the application of this formula can be illustrated by the following example: Suppose the deviation results output by step one for three consecutive frames are 、 、 , and the maximum difference in the short-term sequence is , set the normalization reference value , the trend weight , the structure identifier , then the maximum deviation term is , the second-order difference term is , after normalization, it is , so the denominator is , and we get , indicating that this deviation has a high continuity. When another set of data is 、 、 , the maximum difference is , after normalization, it is , the second-order difference is , after normalization, it is , at this time the denominator is , and we get , indicating that this deviation belongs to the mutation state. If the structure identifier ,but This directly reflects that the current deviation lacks a basis for weld structure analysis. According to the above value range, C=0.57 falls into the range that can be used as the main control basis, indicating that the deviation maintains a good continuity with the short-term deviation sequence; C=0.14 falls into the low confidence range, indicating that the deviation has obvious jumps or sudden changes in trend, and is more likely to correspond to deviations caused by spatter, reflection, obstruction or local anomalies.

[0029] The above calculation process shows that the deviation confidence level The output of step one and The deviation is directly calculated and its continuity and trend are quantitatively evaluated by combining short-time series characteristics. This result is used in subsequent steps to generate control commands, enabling the control system to... Size adjustment The response level is improved, thereby effectively suppressing mistracking caused by spatter, reflection or local anomalies during the welding process and improving the stability of overall welding control.

[0030] S3: Generate candidate control instructions based on the deviation confidence results.

[0031] Specifically, this step follows the deviation confidence result output in step two. And compare it with the weld deviation results at the corresponding time. Used in pairs to generate candidate control commands .in, The weld deviation in the current frame corresponding to the credibility determination in step two is derived from the weld deviation results continuously output in step one. For step two based on The reliability assessment results are derived from weld structure markings and short-term deviation changes. These two factors play different roles in control command generation. Used to determine the direction and target size of the welding torch's lateral correction. This is used to determine the degree to which the current deviation participates in the control. During the welding process, the lateral movement of the welding torch is directly related to the arc position, the stability of the molten pool, and the weld bead formation. When the control quantity jumps abruptly, it can easily cause the arc heat input position to shift rapidly. Therefore, this step introduces both the deviation confidence and the continuity of the control action into the candidate control command generation process, so that the welding torch can both follow the actual weld seam deviation and maintain a smooth movement when the deviation confidence is low or the target changes too much.

[0032] The calculation of candidate control commands is based on the proportional tracking update relationship in discrete control, meaning that the current control quantity gradually approaches the target control quantity along the direction of the actual control command executed in the previous cycle. For welding lateral correction, the target control quantity can be derived from... It means that, among them The proportional coefficient, determined during the welding process debugging phase, is used to convert weld deviation into a welding torch lateral correction target; the actual control command executed in the previous cycle is recorded as follows. The data is recorded by the welding torch actuator controller after the action is completed in the previous control cycle. The classic proportional update relationship can be understood as "the current candidate control command equals the actual executed control command of the previous cycle plus a portion of the target difference." This application adds two constraints suitable for welding scenarios to this relationship: firstly, the output of step two... As the credible participation weight of the current deviation; secondly, add a suppression term formed by the change in the target control quantity to the denominator of the weight, so that the participation ratio of the target control quantity with large jumps is automatically reduced; when the output of step two At that time, this step directly commands This indicates that no new lateral correction increment is introduced in the current control cycle. The candidate control command is obtained as follows: ; in, The candidate control command generated in this step represents the lateral correction amount that is to be output to the welding torch actuator in the current control cycle; The actual control commands executed in the previous cycle are recorded by the welding torch actuator controller. The deviation confidence result is output from step two; The weld deviation result for the current frame is derived from the output of step one and step two. One-to-one correspondence; To control the scaling factor, used to convert weld deviation into welding torch correction target; The maximum permissible variation baseline, configured by the welding torch's lateral execution capability and the process's allowable correction range, is used to normalize variations in the target control quantity. Take greater than If the value is positive, and the corresponding parameter for the current process segment has not yet been updated, then the value corresponding to the previously confirmed process segment will be used. ; is the action mutation suppression coefficient, pre-configured by the controller, used to adjust the degree to which target mutations suppress candidate control commands. Where... and Being in the same control coordinate system, the difference between the two represents the distance between the actual control command executed in the previous cycle and the current target control quantity; The normalized control target change ratio allows the additional term and constant term in the denominator to be combined in the calculation; finally, the right side is obtained by superimposing a control increment on the actual control command executed in the previous cycle to obtain the current candidate control command, and the calculation relationship remains consistent.

[0033] From the perspective of formula derivation, if we take Let it be denoted as the target control variable, then This is the target difference that the controller needs to eliminate in this cycle; if it directly approximates the target by a fixed ratio, a large control increment will be generated when the weld deviation changes abruptly. This step puts the normalized magnitude of the target difference into the denominator. When the target difference is small, the denominator is close to... The controller is mainly based on The value of the denominator is close to the target; when the target difference is large, the denominator increases, and the control increment is compressed. This processing makes the control output jointly determined by the confidence level and the change in action: when the confidence level is high and the target change is gradual, the welding torch quickly approaches the center of the weld; when the confidence level is low or the target change is obvious, the welding torch action tends to be conservative, which is suitable for instantaneous deviation scenarios caused by spatter obstruction, reflective jumps, and local bevel abnormalities in the welding site.

[0034] The specific calculation process is explained below. Let's assume that the control command actually executed in the previous cycle is... Current weld deviation Control proportional coefficient The credibility result output in step two Maximum permissible variation benchmark Action mutation suppression coefficient Then the target control quantity is The target difference is The normalized target change ratio is The trusted participation weight is ,therefore At this point, the welding torch smoothly corrects itself along the center of the weld. In another control cycle, due to localized reflection, the current deviation changes... The actual control command executed in the previous cycle was still... Step 2 outputs the credibility score. If all other parameters remain unchanged, then the target control quantity is The target difference is The normalized target change ratio is The trusted participation weight is ,therefore The results show that even if the single-frame deviation increases, the candidate control command only changes slightly, and the welding torch will not swing drastically due to instantaneous abnormal deviation.

[0035] After the candidate control commands are generated, the controller will This output, taken as the result of this step, is passed to the next step for pre-execution verification. This output incorporates the weld deviation results from step one, the deviation reliability results from step two, and the actual control commands executed in the previous cycle. The next step will... As input, the final execution control command is determined by combining the allowable range of the actuator and the process boundary, so that the candidate control command is further constrained before it actually enters the welding torch actuator.

[0036] S4: Apply execution constraints to the candidate control instructions and issue execution control instructions.

[0037] Specifically, this step receives the candidate control commands output from step three. and convert it into execution control instructions. The signal is then sent to the welding torch actuator, enabling the welding torch to complete the actual lateral correction within the current control cycle. Here... The weld deviation results from step one are already included. The deviation confidence result formed in step two Therefore, the task of this step focuses on the execution level: constraining the candidate control commands according to the motion capability of the welding torch actuator and the continuity of motion allowed by the welding process, and then actually executing them through the robot controller, welding torch slide driver, or servo control interface. During welding, the lateral movement of the welding torch directly changes the arc's position. If the movement is too fast within a single control cycle, it can easily cause a sudden shift in the arc's point of action, resulting in fluctuations at the edge of the molten pool; if the movement is too slow, the welding torch will lag behind the actual weld center. Therefore, this step adopts an execution speed limit constraint to transform the candidate control commands obtained in step three into execution control commands that conform to the equipment's motion capability.

[0038] The calculation of the execution control command originates from the limiting update relationship in industrial servo control, that is, limiting the difference between the target command and the command executed in the previous cycle to within the allowable range of variation in a single cycle. Let the actual execution control command issued in the previous cycle be... This value is recorded by the welding torch actuator controller after the previous cycle has completed its action; the candidate control command input in step three for this cycle is... The allowable range of change per single period is Its value is configured based on the maximum response capability of the welding torch's lateral actuator, the welding speed, and the allowable smoothness of the process. Based on the above constraints, the execution control command is determined by the following formula: ; in, The execution control command output in this step represents the lateral correction amount actually issued to the welding torch actuator in the current cycle; These are the candidate control commands output from step three; The horizontal correction amount that was issued and completed by the actuator in the previous cycle is recorded by the actuator controller; The allowable variation range per cycle is set by the capability of the welding torch's lateral actuator and the welding process parameters. The logic of this formula is: when the change in the candidate control command relative to the previous executed control command exceeds the positive allowable range, only the change in the current cycle is increased. When the change is within the allowable range, the candidate control instruction is executed directly; when the change exceeds the reverse allowable range, the control instruction is reduced only in this cycle. This formula belongs to the constraint projection processing of candidate instructions, so that the final issued instruction is and Maintaining the same control coordinate system and ensuring that the welding torch movement changes within each control cycle remain within the tolerance range of the actuator; if obtained according to the above formula... If the lateral correction exceeds the allowable range for the current weld segment, then... The corresponding boundary value is captured, and the amplitude limit flag is written into the control record.

[0039] Taking a welding process as an example, if the control command was executed in the previous cycle... Candidate control commands output in step three Permissible variation range per single period ,but It is within the allowable range of change, therefore The welding torch directly performs lateral correction according to the candidate control command. If, in another cycle, the weld deviation suddenly increases, step three outputs... The control command executed in the previous cycle was still... ,but The change exceeds the allowable range for a single cycle, therefore the control command executed in this cycle is... The welding torch first completes a smooth movement within a limited range. If the candidate control command changes in the opposite direction, for example... , , ,but The control command executed in this cycle is: This calculation process enables candidate control commands to be actually input into the welding torch actuator, while ensuring that the executed actions meet the welding process's requirements for arc stability and weld pool continuity.

[0040] The controller will The control signal is sent to the welding torch actuator via the welding robot control interface or servo drive interface, and the actuator follows... Adjust the horizontal position of the welding torch, and after completing the action... The previous control command is recorded as the starting control command for the next control cycle. If no position feedback, servo response, or robot controller confirmation signal is received within the preset feedback time, the controller retains the previously confirmed control command as the starting execution quantity for the next control cycle and writes an execution anomaly flag. As the welding torch moves, the laser tracking module re-acquires the weld contour in the next sampling cycle. Step one generates a new weld deviation result, step two generates a new deviation confidence result, and step three generates new candidate control commands. Thus, the current step completes the conversion from candidate control commands to actual welding torch actions, and the execution results of each cycle are naturally connected to the weld tracking control process of the next cycle. In system implementation, the above method can be completed collaboratively by a laser contour acquisition unit, a bevel structure recognition unit, a deviation confidence determination unit, a candidate control command generation unit, and an execution control command constraint issuance unit to form a corresponding welding control system.

[0041] In one or more embodiments, such as Figure 2 As shown, a laser-tracking-based welding control system is disclosed, the system comprising: Laser contour acquisition unit, configured to acquire weld contour data; A bevel structure recognition unit is configured to extract weld deviation results. A deviation reliability determination unit is configured to determine the reliability of the weld deviation result and form a deviation reliability result. A candidate control instruction generation unit is configured to generate candidate control instructions based on the deviation confidence result.

[0042] It is worth noting that the specific workflow of the laser-tracking-based welding control system provided in this embodiment of the invention is the same as that of the laser-tracking-based welding control method described in the above embodiment, and will not be repeated here.

[0043] This invention also provides a laser-tracking-based welding control device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the steps described in the above embodiment of a laser-tracking-based welding control method, for example... Figure 1 The steps S1 to S4 described above; or, when the processor executes the computer program, it implements the functions of each module in the above system embodiments.

[0044] For example, the computer program may be divided into one or more modules, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules may be a series of computer program instruction segments capable of performing specific functions, which describe the execution process of the computer program in the laser-tracking-based welding control device.

[0045] The laser-tracking-based welding control device can be a desktop computer, laptop, handheld computer, or cloud server, or other computing device. This laser-tracking-based welding control device may include, but is not limited to, a processor and memory. Those skilled in the art will understand that the laser-tracking-based welding control device may also include input / output devices, network access devices, buses, etc.

[0046] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the laser-tracking-based welding control equipment, connecting all parts of the equipment via various interfaces and lines.

[0047] The memory can be used to store the computer program and / or modules. The processor implements various functions of the laser-tracking-based welding control device by running or executing the computer program and / or modules stored in the memory, and by calling the data stored in the memory. The memory may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function, etc.; the data storage area may store data created according to the operation of the controller, etc. In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital card (SD card), flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0048] The integrated module of the laser-tracking-based welding control equipment, if implemented as a software functional unit and sold or used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0049] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0050] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A laser tracker based welding control method, characterized by, The method includes: Acquire weld contour data and extract weld deviation results; The reliability of the weld deviation results is determined to form a deviation confidence result; the steps of determining the reliability of the weld deviation results include: A short-time deviation sequence is introduced to compare the weld deviation of the current frame with its historical changes; The variation characteristics of the deviation sequence are described based on the maximum deviation amplitude and discrete second-order difference, and the deviation confidence result is calculated by combining the weld structure identification. The steps for calculating the deviation confidence result based on the maximum deviation amplitude and discrete second-order difference to describe the variation characteristics of the deviation sequence, and combined with weld structure identification, include: Calculate the maximum deviation magnitude of the weld deviation in the current frame relative to the weld deviation in historical frames; Calculate the discrete second-order difference between the weld deviation in the current frame and the weld deviation in the historical frames; The maximum deviation amplitude and discrete second-order difference are fused with the weld structure identifier to obtain the deviation confidence result, wherein the weld structure identifier is determined based on the groove boundary spacing and the continuity of the contour change. Candidate control instructions are generated based on the deviation confidence results; The candidate control instructions are subject to execution constraints, and the execution control instructions are issued.

2. The laser tracker-based welding control method of claim 1, wherein, The steps of acquiring weld contour data and extracting weld deviation results include: The weld contour data is obtained by the laser tracking module. The weld search window is captured with the reference center position of the welding gun. The bevel structure is found in the window, and then the left bevel boundary and the right bevel boundary are identified. The midpoint between the left and right bevel boundaries is calculated as the geometric center of the weld. The geometric center of the weld is then compared with the reference center of the welding torch to obtain the weld deviation result.

3. The laser tracker-based welding control method of claim 2, wherein, The steps for identifying the left and right bevel boundaries include: The position where the profile transitions from the flat area of ​​the parent material into the continuously descending area is defined as the left bevel boundary. The right bevel boundary is defined as the point where the contour rises continuously from the bevel area and then flattens out again. If multiple sets of candidate left and right boundaries that meet the conditions are formed in the current search window, the set whose midpoint is closest to the geometric center of the weld confirmed in the previous control cycle shall be selected as the left bevel boundary and the right bevel boundary.

4. The laser tracker-based welding control method of claim 1, wherein, The step of generating candidate control instructions based on the deviation confidence result includes: The deviation confidence result is used as the confidence participation weight of the current deviation; The deviation confidence result is combined with the weld deviation result of the current frame to generate candidate control commands, so that the welding torch maintains smooth operation when the deviation confidence is low or the target changes too much.

5. The laser-tracking-based welding control method according to claim 4, characterized in that, The step of using the deviation confidence result as the confidence participation weight of the current deviation includes: The target control quantity is calculated by combining the deviation confidence result with the weld deviation result of the current frame. The difference between the target control quantity and the actual control command executed in the previous cycle is calculated as the target difference. The normalized magnitude of the target difference is placed in the denominator. When the target difference is large, the denominator value is increased to compress the control increment.

6. The laser-tracking-based welding control method according to claim 1, characterized in that, The step of imposing execution constraints on the candidate control instructions and issuing execution control instructions includes: The candidate control instruction is compared with the execution control instruction actually issued in the previous cycle; When the change range of the candidate control instruction exceeds the allowable change range in a single cycle, the candidate control instruction is truncated as the sum of the execution control instruction actually issued in the previous cycle and the allowable change range in a single cycle, and an execution control instruction is formed.

7. The laser-tracking-based welding control method according to claim 1, characterized in that, After the step of imposing execution constraints on the candidate control instructions and issuing execution control instructions, the method further includes: The execution control command after the welding torch actuator completes its action is recorded as the previous execution control command for the next control cycle; If no position feedback or servo response is received within the preset feedback time, the previously confirmed executed control command is retained as the starting execution value for the next control cycle.

8. A laser-tracking-based welding control system, characterized in that, The system includes: Laser contour acquisition unit, configured to acquire weld contour data; A bevel structure recognition unit is configured to extract weld deviation results. A deviation reliability determination unit is configured to determine the reliability of the weld deviation result and generate a deviation reliability result; the step of determining the reliability of the weld deviation result includes: A short-time deviation sequence is introduced to compare the weld deviation of the current frame with its historical changes; The variation characteristics of the deviation sequence are described based on the maximum deviation amplitude and discrete second-order difference, and the deviation confidence result is calculated by combining the weld structure identification. The steps for calculating the deviation confidence result based on the maximum deviation amplitude and discrete second-order difference to describe the variation characteristics of the deviation sequence, and combined with weld structure identification, include: Calculate the maximum deviation magnitude of the weld deviation in the current frame relative to the weld deviation in historical frames; Calculate the discrete second-order difference between the weld deviation in the current frame and the weld deviation in the historical frames; The maximum deviation amplitude and discrete second-order difference are fused with the weld structure identifier to obtain the deviation confidence result, wherein the weld structure identifier is determined based on the groove boundary spacing and the continuity of the contour change. A candidate control instruction generation unit is configured to generate candidate control instructions based on the deviation confidence result.

Images