Semi-automatic manual welding high-efficiency track-finding auxiliary system

By combining micro laser triangulation and infrared thermal imaging, the wire feeding speed and contact tip displacement are adjusted in real time, the welder's intentions are identified, and multi-level safety protection is achieved. This solves the problems of track-finding stability and safety of semi-automatic welding systems under complex working conditions, and improves welding quality and efficiency.

CN123103121APending Publication Date: 2026-11-17HANGXIAO STEEL STRUCTURE (TANGSHAN) CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611632936.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-10-21
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing semi-automatic welding systems face problems such as severe sensor data interference, logic jumps, limited response speed, insufficient flexibility, difficulty in recognizing welder's intentions, and insufficient safety under complex working conditions, leading to malfunctions such as unstable track tracking, welding wire adhesion, or burnout.

Method used

By combining a miniature laser triangulation sensor and infrared thermal imaging, the distance between the welding wire and the bevel and the information of the molten pool are obtained in real time. By adjusting the wire feeding speed and the displacement of the contact tip, the arc force is used to correct the deviation. An integrated pressure sensor identifies the welder's intention and monitors the current and voltage parameters in real time to achieve multi-level safety protection.

Benefits of technology

It improves the accuracy and reliability of track-finding data, enhances the system's dynamic tracking capability under complex trajectories, enables human-machine collaborative operation, ensures the continuity and quality consistency of the welding process, and prevents welding wire from sticking and burning.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123103121A_ABST
    Figure CN123103121A_ABST
Patent Text Reader

Abstract

This invention belongs to the field of welding auxiliary technology, specifically disclosing a semi-automatic manual welding high-efficiency track-finding auxiliary system. The system includes: a data acquisition module that uses a miniature laser triangulation sensor to obtain the real-time distance to the welding wire tip; a distance deviation calculation module that calculates the deviation vector value; an adjustment module that adjusts the welding wire extension length by coordinating changes in wire feed speed and contact tip displacement; a pointing control module that uses the physical characteristics of an electric arc to achieve flexible correction; a monitoring module that combines infrared thermal distribution images for status determination; and an analysis module that identifies the welder's intentions and adjusts priorities using a pressure sensor array. This application achieves non-contact trajectory-assisted correction and deep human-machine collaboration through dual laser and infrared sensing and microsecond-level response adjustment of the welding wire length. While overcoming interference at the welding site, it significantly improves the reliability of track finding, dynamic tracking capability, and system operational safety.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of welding auxiliary technology and relates to a semi-automatic manual welding high-efficiency track-finding auxiliary system. Background Technology

[0002] In modern industrial manufacturing, welding technology, as a core process for connecting metal structures, directly impacts product manufacturing quality and production efficiency through its level of automation and intelligence. With the increasing demands for consistency in welded joints in high-end equipment manufacturing, intelligent assisted welding systems with sensing and adaptability have become a crucial direction for industry development. Real-time tracking and precise positioning during the welding process are key foundations for achieving high-quality welding. Especially in the assembly of complex structural components, efficient track-finding technology can significantly reduce over-reliance on operator skills and experience, ensuring the stability of weld quality.

[0003] Semi-automatic manual welding track-tracking systems integrate sensors to detect weld bevels in real time and use control logic to assist in correcting the welding trajectory. This demonstrates significant application value in non-standard parts processing, welds with varying curvature, and confined working spaces. The core of this technology lies in establishing a spatial mapping relationship between the sensor reference, the welding wire tip, and the workpiece bevel. By dynamically adjusting process parameters or actuators, the system guides the welding wire tip to always precisely align with the bevel centerline. An effective track-tracking system not only requires an extremely high data acquisition frequency but also the ability to handle the complex physical environment of the welding site to achieve microsecond-level perception and precise feedback of the welding trajectory.

[0004] Existing technologies for semi-automatic welding track finding often face challenges from multiple complex working conditions, including intense arc light, molten metal spatter, and welder hand tremors. This results in severe transient interference and logic jumps in the raw data collected by sensors. Furthermore, traditional correction mechanisms typically rely on single mechanical transmission adjustments, limiting response speed and lacking in-depth utilization of the physical characteristics of the arc, making it difficult to achieve flexible restoring force guidance during high-speed welding. In addition, the lack of accurate recognition of the welder's intention to intervene easily leads to logical conflicts between automatic control logic and manual operation commands. Traditional systems have weak comprehensive monitoring capabilities for molten pool heat distribution characteristics, wire resistance heat accumulation, and fluctuations in electrophysical parameters. In cases of abrupt changes in bevel shape or sensor failure, technical malfunctions such as track finding instability, wire adhesion, or even burnout are highly likely to occur. Summary of the Invention

[0005] In view of this, in order to solve the problems mentioned in the background technology, a semi-automatic manual welding high-efficiency track-finding auxiliary system is proposed.

[0006] The objective of this invention can be achieved through the following technical solution: This invention provides a semi-automatic manual welding high-efficiency track-finding auxiliary system, including: an environmental data acquisition module, which uses a miniature laser triangulation distance sensor to obtain the real-time physical distance between the end of the welding wire and the bottom of the bevel on the workpiece surface.

[0007] The distance deviation calculation module compares the real-time physical distance with a preset reference distance benchmark and calculates the deviation vector value.

[0008] The wire extension adjustment module, based on the deviation vector value, dynamically adjusts the wire extension length by changing the wire feeding speed and the physical displacement of the contact tip in the axial direction.

[0009] The directional correction control module utilizes the effect of the change in the directionality of the arc force caused by the change in the extension length of the welding wire to generate a lateral restoring force at the end of the welding wire that points towards the center line of the bevel.

[0010] The track-seeking status monitoring module acquires infrared thermal distribution images of the molten pool and the rear area, obtains the centroid coordinates of the molten pool, compares them with the expected center position calculated through distance information, and determines the validity of the current track-seeking status.

[0011] The control priority analysis module collects lateral pressure data applied by the welder through a pressure sensor array, determines the welder's operating intention, and adjusts the control priority of automatic track finding accordingly.

[0012] The closed-loop safety protection module monitors the current and voltage parameters in the welding circuit in real time. When abnormal fluctuations or exceeding the safe range are detected, an alarm signal is issued and the drive mechanism is locked.

[0013] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. The present invention overcomes the interference of strong light and spatter at the welding site by using dual sensing of micro laser triangulation and infrared thermal imaging, thereby improving the accuracy and reliability of track tracking data.

[0014] This invention combines wire feeding speed adjustment with conductive tip axial displacement adjustment to achieve microsecond-level response to the wire extension length, thereby enhancing the system's dynamic tracking capability under complex trajectories.

[0015] This invention delves into the physical characteristics of electric arcs and utilizes the lateral restoring force generated by the change in welding wire length to achieve flexible correction, avoiding the problems of bulky structure and insufficient flexibility caused by traditional mechanical correction.

[0016] This invention identifies the welder's intentions by integrating a pressure sensor array, thus resolving the logical conflict between automatic control and manual operation and realizing human-machine collaborative work.

[0017] This invention ensures the safe operation of the system under extreme conditions through a multi-layered safety protection mechanism and feedforward compensation logic, prevents welding wire adhesion and burn-out accidents, and guarantees the continuity of the welding process and the consistency of workpiece quality. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a schematic diagram of the system module connections of the present invention.

[0020] Figure 2 This is a schematic diagram illustrating the flexible correction steps in an embodiment of the present invention.

[0021] Figure 3 This is a schematic diagram of the main stages of track-seeking status monitoring in an embodiment of the present invention.

[0022] Figure 4 This is a schematic diagram of the logic flow for identifying the welder's operational intent and adjusting the priority of automatic track-following control in an embodiment of the present invention. Detailed Implementation

[0023] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] Please see Figure 1 As shown, this invention provides a semi-automatic manual welding high-efficiency track-finding auxiliary system, including an environmental data acquisition module, a distance deviation calculation module, a welding wire extension adjustment module, a pointing correction control module, a track-finding status monitoring module, a control priority analysis module, and a closed-loop safety protection module. The environmental data acquisition module and the track-finding status monitoring module are both connected to the distance deviation calculation module. The distance deviation calculation module, the control priority analysis module, and the closed-loop safety protection module are all connected to the welding wire extension adjustment module. The welding wire extension adjustment module is connected to the pointing correction control module.

[0025] The environmental data acquisition module is used to obtain the real-time physical distance between the end of the welding wire and the bottom of the bevel on the workpiece surface using a miniature laser triangulation sensor.

[0026] In a preferred embodiment of the present invention, the real-time physical distance is obtained as follows: the total physical distance from the sensor reference surface to the workpiece surface is collected using a miniature laser triangulation rangefinder, and the fixed physical deviation from the sensor reference surface to the end of the welding wire is subtracted to obtain the real-time physical distance value from the end of the welding wire to the bottom of the bevel.

[0027] It is important to note that obtaining the real-time physical distance from the welding wire tip to the bottom of the workpiece bevel is crucial for the system to determine whether the welding wire deviates from the bevel centerline, whether it is too close or too far, and then calculate the deviation vector to provide a basis for subsequent adjustments. This distance data filters out interference from the electric arc and spatter, ensuring stable measurement. Simultaneously, it is combined with infrared molten pool monitoring for dual verification to prevent tracking failure. Relying on real-time distance feedback, the system can quickly adjust the welding wire extension length and utilize arc force to achieve flexible correction, ensuring accurate welding trajectory and stable molten pool, thereby improving welding quality and operational safety.

[0028] It should be added that the laser generator inside the miniature laser triangulation rangefinder emits a probe laser in a specific red light band, and the optical axis of the probe laser is spatially collinear with the central axis of the welding wire. The probe laser is projected onto the workpiece surface at the bottom of the bevel to form a characteristic light spot. The photosensitive unit inside the sensor captures the diffuse reflection signal generated by the characteristic light spot and converts the received light signal into an analog electrical signal. The analog electrical signal is then processed by analog-to-digital conversion and input to the controller, which uses a preset geometric trigonometric ratio model.

[0029] During the measurement process, the controller initiates high-frequency sampling logic to perform weighted averaging of distance values ​​within multiple consecutive sampling periods. The sampling points with time nodes closer to the current moment are assigned higher weight coefficients. The moving average algorithm filters out instantaneous interference jumps caused by arc fluctuations or metal droplet splashes during the welding process.

[0030] It should be noted that, specifically, sampling points closer to the current time are assigned a higher weight coefficient. For example, the weight coefficient of the current sampling point is set to 0.3, while the weights of previous sampling points decay exponentially. The controller executes a moving average algorithm, continuously calculating the standard deviation between the current sample value and the mean of the previous 50 historical values ​​to monitor the dispersion of the data in real time. When a momentary sampling value is detected to deviate from the mean by more than a preset interference judgment threshold, the system automatically identifies it as a momentary interference jump caused by violent arc fluctuations, molten metal droplet splashes, or smoke obstruction during the welding process, and immediately removes the anomaly. This results in the output of a smooth distance data stream after second-order low-pass filtering, ensuring that the distance parameters fed back to subsequent logic have extremely high physical confidence.

[0031] It should be noted that the interference judgment threshold is set based on the actual interference characteristics at the welding site and the system stability requirements. The core basis is the amplitude of typical instantaneous interferences such as strong arc light, molten metal droplet splashes, fumes, and bevel oxide obstruction. This is combined with the moving average and standard deviation of high-frequency sampled distance data, and matched to different plate materials, bevel types, and welding process parameters. Simultaneously, the distinction between normal trajectory deviations and abnormal interference is considered, ensuring that effective deviations are not misjudged as interference, and that real interference is not missed. This provides a judgment standard for filtering out abnormal points and outputting stable and reliable distance measurement data.

[0032] For example, taking the welding of a V-groove and an 8mm thick low-carbon steel plate as an implementation example, the interference judgment threshold is set based on the actual measured interference amplitude on site: First, laser distance data during normal welding without deviation is collected, and the sliding mean and standard deviation of 50 points are calculated. The threshold is set to the mean ± 3 times the standard deviation, corresponding to the typical jump amplitude caused by arc fluctuation, spatter, and smoke. This threshold can effectively filter out instantaneous interference, retain the true trajectory deviation signal, ensure that the system stably identifies groove offset, and prevents misjudgment and missed detection, providing a reliable basis for filtering and track-following control.

[0033] The distance deviation calculation module is used to compare the real-time physical distance with a preset reference distance benchmark and calculate the deviation vector value.

[0034] In a preferred embodiment of the present invention, the specific calculation method of the deviation vector value is as follows: a reference distance benchmark database for plates of different thicknesses and different bevel forms is pre-constructed and stored.

[0035] It should be explained that the reference distance database is a pre-established standard distance reference library categorized and stored by welding conditions. It is measured and verified under standard working conditions for different plate thicknesses and different bevel types to obtain ideal stable distance values ​​from the welding wire tip to the bottom of the bevel, and these values ​​are organized and saved according to parameter dimensions. Before welding, the system automatically matches the corresponding reference based on the actual plate material and bevel, serving as the standard reference for calculating trajectory deviations.

[0036] It should be noted that the reference distance database needs to be categorized by plate thickness, bevel type, bevel angle, and initial wire extension length. Under controlled welding conditions, multiple sets of calibration welds are completed using standard processes for common bevels such as V-type, U-type, and I-type, and plates of different thicknesses. The stable distance from the tip of the welding wire to the bottom of the bevel is measured by a laser sensor. After filtering and repeated verification, the effective reference values ​​are extracted, organized and stored in the database according to parameter dimensions, forming a reference data set that can be automatically matched and called, providing a standard reference for deviation calculation.

[0037] Before the welding operation begins, the corresponding reference distance reference value is automatically matched in the reference distance reference database based on the input plate thickness, bevel angle and preset initial extension length of welding wire.

[0038] Once the welding process begins, the controller extracts the filtered real-time physical distance at fixed time intervals and subtracts the reference distance baseline value to obtain the difference.

[0039] If the difference is positive, it is determined that the physical position of the welding wire tip relative to the bottom of the groove has been raised, and it is identified as the trajectory of the manual welding gun deviating from the center line of the groove and shifting towards the side wall.

[0040] If the difference is negative, it is determined that the end of the welding wire has sunk, which is identified as the end of the welding wire being too close to the surface of the workpiece.

[0041] The controller defines the directionality of the deviation based on the positive or negative attribute of the difference, and defines the intensity attribute of the deviation based on the absolute value of the difference, thus synthesizing a deviation vector with both directionality and intensity attributes.

[0042] It should be noted that the controller defines the direction attribute of the deviation based on the polarity of the difference, with a positive sign indicating distance from the root and a negative sign indicating proximity to the root. The controller also defines the intensity attribute of the deviation based on the absolute value of the difference. Finally, the system encapsulates the direction and intensity attributes into a deviation vector containing amplitude and phase, which serves as the sole control input for subsequent actuator drive logic.

[0043] The wire extension adjustment module is used to dynamically adjust the wire extension length based on the deviation vector value by changing the wire feeding speed and the physical displacement of the contact tip in the axial direction.

[0044] It should be noted that the purpose of adjusting the wire extension length is to correct the distance deviation between the wire tip and the bottom of the bevel in real time, ensuring that the wire is always aligned with the center line of the bevel for stable welding. This adjustment can counteract trajectory deviations caused by the arc, spatter, and hand tremors. It utilizes changes in arc force to create a flexible correction effect, preventing the wire from being too high (leading to arc breakage) or too low (causing adhesion). Simultaneously, it ensures stable molten pool shape and welding energy, improves trajectory tracking accuracy and welding quality, and guarantees a continuous, safe, and reliable welding process.

[0045] In a preferred embodiment of the present invention, the specific method for achieving dynamic adjustment of the welding wire extension length is as follows: a collaborative conversion model between the wire feeding speed and the conductive tip displacement is established inside the controller.

[0046] It should be noted that this collaborative conversion model can quantify the linkage between wire feed speed and contact tip displacement into a unified control command, achieving dual-parameter coupled adjustment. Its core advantage lies in its ability to quickly compensate for welder hand tremors and minor errors in workpiece bevel, converting the positional deviation of the welding wire tip relative to the bevel into precise motion compensation. This model can dynamically match the optimal extension length at different welding speeds, ensuring arc stability while improving tracking response speed and anti-interference capabilities, ensuring welding trajectory consistency, and significantly enhancing welding quality and automation levels.

[0047] When the deviation vector value indicates that the wire extension length needs to be increased, the controller outputs a first command to the wire feeding drive mechanism to reduce the pulse frequency of the wire feeding motor to slow down the rate at which the wire is fed into the molten pool. At the same time, it outputs a second command to the conductive tip axial drive mechanism to drive the micro stepper motor to retract the conductive tip in a direction away from the workpiece through a precision lead screw pair.

[0048] It should be added that during this adjustment process, the controller performs quantitative calculations based on the amplitude and direction of the deviation vector value, combined with the built-in wire feeding speed and conductive tip displacement collaborative conversion model: first, the target increment of the wire extension length is determined according to the deviation intensity, and then the increment is decomposed into the reduction value of the wire feeding motor pulse frequency and the displacement of the conductive tip according to the calibration ratio. The two are output synchronously and matched in time, and precise displacement is achieved through the micro stepper motor and lead screw pair, and finally the quantitative adjustment is completed.

[0049] Furthermore, the calibration ratio is a quantitative correspondence between the pulse frequency of the wire feed motor and the axial displacement of the contact tip, used to ensure precise synchronization of their coordinated adjustment. It is achieved through offline standard experiments, where the pulse frequency is changed and the contact tip displacement is measured under a fixed welding torch state. Multiple sets of data are collected to fit linear coefficients, which are then corrected by combining the bevel type, welding wire specifications, and welding process parameters before being written into the controller. This becomes the core conversion basis for the system to calculate the pulse frequency reduction and the contact tip retraction distance.

[0050] It should be noted that by actively slowing down the wire feeding speed and actively retracting the position of the contact tip, a rapid and dynamic increase in the wire extension length is achieved in terms of macroscopic physical morphology.

[0051] When the deviation vector value indicates that the wire extension length needs to be reduced, the controller outputs a third command to increase the pulse frequency of the wire feeding motor to accelerate the wire replenishment, and at the same time outputs a fourth command to drive the micro stepper motor to extend the conductive tip forward toward the workpiece.

[0052] It should be noted that the controller first calculates the target value that the wire extension length needs to be reduced based on the amplitude and direction of the deviation vector. Then, it performs a quantitative conversion according to the preset wire feeding pulse frequency and the conductive tip displacement calibration ratio to determine the specific pulse frequency value that the wire feeding motor needs to increase. Simultaneously, it calculates the precise displacement of the conductive tip toward the workpiece and converts the two into directly executable third and fourth control commands to achieve precise quantitative adjustment of the wire extension length.

[0053] It should be noted that the collaborative conversion model ensures that the rate of change of wire feed speed and the rate of change of contact tip displacement remain highly consistent over time, achieving stepless smooth adjustment of the wire extension length within a preset physical stroke range without changing the welding energy input and the heat balance of the molten pool.

[0054] Furthermore, this dual-physical-dimensional regulation mechanism can significantly counteract hand tremors during manual operation, greatly shorten the closed-loop response time of the system to trajectory deviations, and ensure the stability of the droplet transfer pattern during welding.

[0055] The pointing correction control module is used to generate a lateral restoring force pointing towards the center line of the bevel at the end of the welding wire by utilizing the change in the directionality of the arc force caused by the change in the extension length of the welding wire.

[0056] For a preferred embodiment of the present invention, please refer to Figure 2 The specific logic for generating the transverse restoring force is as follows: based on the positive correlation between the wire extension length and resistance heat in gas metal arc welding, when the wire extension length increases, the resistance heat component of the wire itself increases, thereby changing the plasma hydrodynamic distribution of the arc at the end of the wire.

[0057] It needs to be explained that as the extension length of the welding wire increases, the path through which the current flows through the welding wire becomes longer. According to Joule's law, the resistive heat component of the welding wire itself increases significantly, which directly leads to an increase in the enthalpy of the metal at the end of the welding wire, thereby changing the plasma hydrodynamic distribution state of the arc at the end of the welding wire.

[0058] When the wire extension length exceeds the preset process stability value, the axial thrust component generated by the arc weakens, while the transverse force component acting on the bevel sidewall is enhanced due to the expansion of the plasma column, resulting in a physical tendency to push the wire end towards the bevel sidewall.

[0059] It should be explained that the stability value of this process is based on the core settings of welding wire specifications, welding current and voltage, groove type and metal melting characteristics. Combined with offline experimental measurements of arc force distribution, droplet transition state and welding stability data under different extension lengths, the critical length value for balancing arc axial thrust and transverse recovery force and welding process without arc interruption and spatter is determined. At the same time, the adjustment range of wire feed and contact tip and the safe range of molten pool heat input are matched to ensure reliable correction and stable operation.

[0060] When the wire extension length is less than the preset process stability value, the arc energy concentrates in the center area at the bottom of the bevel, generating an attractive restoring force pointing towards the root of the bevel.

[0061] By finely adjusting the wire extension length using a controller, the spatial distribution change of the arc force is transformed into a restoring torque pointing towards the center line of the bevel.

[0062] In a preferred embodiment of the present invention, the method further includes a pre-calibration process for the transfer function of distance and restoring force, the specific steps of which are as follows: for V-shaped bevels, U-shaped bevels and I-shaped bevels, an independent experimental calibration database is established in advance under a controlled environment.

[0063] During the calibration process, by fixing the manual welding torch and manually changing the wire extension length, a three-dimensional force sensor is used to measure the numerical changes of the transverse arc force on the wire end under different bevel geometries, forming a multi-dimensional lookup table model.

[0064] The lookup table model uses the bevel angle, bevel depth, real-time welding current, real-time welding voltage, and current wire extension length as multi-dimensional input variables, and the expected lateral restoring force as the output variable.

[0065] During actual online welding, the controller calls the corresponding lookup table model in real time based on the identified bevel type, and calculates the precise adjustment amount of the wire extension length required under the current instantaneous working condition through interpolation algorithm.

[0066] The track-seeking status monitoring module is used to collect infrared thermal distribution images of the molten pool and the rear area, obtain the centroid coordinates of the molten pool, compare them with the expected center position calculated through distance information, and determine the validity of the current track-seeking status.

[0067] It should be explained that the purpose of determining the validity of the current tracking status is to verify the authenticity and reliability of the laser ranging data, and to avoid ranging distortion and automatic correction malfunctions caused by bevel oxides, impurities, or strong light interference. By comparing the infrared molten pool centroid with the theoretical bevel center, tracking failures can be identified in a timely manner. When the deviation exceeds the limit, automatic adjustment is paused and parameters are locked to prevent problems such as welding wire deviation, molten pool abnormalities, and welding defects, ensuring the stability, accuracy, and safety of the entire tracking and correction process.

[0068] For a preferred embodiment of the present invention, please refer to Figure 3 As shown, the specific method for determining the validity of the current tracking state is as follows: the infrared imaging component eliminates visible light interference through a long-wave infrared filter and aligns its field of view center with the liquid molten pool area behind the welding arc.

[0069] The controller acquires raw infrared data frames containing the temperature field distribution of the molten pool, performs grayscale processing on the raw infrared data frames, and uses a preset grayscale threshold to segment the image into regions, identifying continuous regions with grayscale values ​​higher than the high temperature threshold as the target areas of the molten pool.

[0070] Preferably, an adaptive grayscale threshold setting method is adopted. First, grayscale statistics and Gaussian filtering preprocessing are performed on the real-time image of the infrared molten pool. The peaks and valleys of the grayscale histogram of the entire image are calculated. The valley value at the grayscale boundary between the molten pool area and the background area is selected as the initial threshold. Then, the threshold is dynamically fine-tuned in combination with the intensity of on-site smoke and splash interference, so that the threshold can stably distinguish the effective area of ​​the molten pool from the interference noise, ensuring that the centroid and contour extraction of the molten pool are accurate and reliable.

[0071] Calculate the arithmetic mean of the horizontal and vertical coordinates of all pixels within the target area of ​​the molten pool, and define it as the centroid coordinate of the molten pool.

[0072] The controller extracts the projection position of the centroid coordinates of the molten pool onto the transverse axis and compares it in real time with the theoretical bevel center coordinates calculated by the micro laser triangulation sensor through geometric transformation.

[0073] If the spatial deviation between the two values ​​exceeds the preset safety deviation limit, the controller determines that the current laser ranging data is obstructed by oxides or impurities on the bevel surface, causing abnormal ranging. Subsequently, it outputs a command to forcibly pause the automatic adjustment logic and lock the current wire feeding parameters until the infrared imaging component confirms that the centroid coordinates of the molten pool have returned to the logic allowable range.

[0074] It should be noted that the safety deviation limit is set through a combination of offline welding tests and process parameters. First, for different plate thicknesses, bevel types, and welding current and voltage, the critical deviation values ​​of normal wire tracking and abnormal deviation are measured. Then, the basic limit is determined by combining the stable range of the molten pool and the allowable oscillation amplitude of the welding wire, while reserving a certain safety margin to avoid misjudgment due to slight vibrations. Finally, this limit is written into the controller as the basis for determining whether the tracking is effective and whether to activate correction or alarm.

[0075] The control priority analysis module is used to collect lateral pressure data applied by the welder through a pressure sensor array, determine the welder's operating intention, and adjust the control priority of automatic track finding accordingly.

[0076] For a preferred embodiment of the present invention, please refer to Figure 4As shown, the specific method for determining the welder's operating intention is as follows: multiple thin-film pressure gauges are embedded in the left and right gripping areas of the manual welding gun handle to form a pressure sensor array.

[0077] The controller collects the analog voltage signals of each diaphragm pressure gauge in real time, and calculates the pressure difference between the total pressure on the left and the total pressure on the right by weighted summation, which is recorded as the lateral operating pressure deviation.

[0078] The lateral operating pressure deviation signal is fed into a digital low-pass filter with a preset cutoff frequency for smoothing, eliminating high-frequency components generated by the physiological tremors of the welder's hands.

[0079] The filtered pressure difference is compared with a preset intervention threshold. If the absolute value of the pressure difference is greater than the intervention threshold, it is determined that the welder is performing an active trajectory correction operation. The controller automatically reduces the gain coefficient of the automatic track-finding control loop to put the automatic adjustment system in a follow-up state.

[0080] It should be noted that the intervention judgment threshold is set comprehensively based on the normal operating pressure fluctuation range of the pipeline and typical interference characteristics. First, multiple sets of pressure difference data are collected when the system is running stably, and their mean and standard deviation are calculated. Then, the threshold range is determined according to the 3σ principle. At the same time, it is corrected by combining the characteristics of the pipeline medium, flow changes and environmental disturbance amplitude. Finally, a critical value that can effectively distinguish between normal fluctuations and abnormal interference is set to ensure that the system accurately identifies abnormalities and triggers corresponding intervention actions.

[0081] It should be added that the control priority analysis module further includes a handle tilt angle compensation function. By integrating a high-precision triaxial accelerometer inside the manual welding torch handle, the controller can acquire the tilt angle information of the welding torch in three-dimensional space in real time. Based on the spatial attitude information, the controller performs gravity vector decomposition and dynamically compensates for the gravity component of the lateral operating pressure deviation collected by the pressure sensor array. This process can eliminate the additional squeezing force on the diaphragm pressure gauge caused by the weight of the welding torch itself at different welding positions, so that the pressure discrimination logic can purely reflect the subjective operating force applied by the welder to the handle, avoiding misjudgments caused by changes in welding posture.

[0082] The closed-loop safety protection module is used to monitor the current and voltage parameters in the welding circuit in real time. When abnormal fluctuations or exceeding the safe range are detected, an alarm signal is issued and the drive mechanism is locked.

[0083] In a preferred embodiment of the present invention, the specific analysis process of the closed-loop safety protection module is as follows: the electrical physical parameters of the welding power supply output terminal are extracted in real time through the Hall current sensor and the voltage sampling circuit, and the product of the instantaneous value of the welding current and the instantaneous value of the welding voltage is calculated and recorded as the real-time energy input value of welding.

[0084] Using a preset resistance thermophysical model of the welding wire, combined with the real-time monitored value of the welding wire extension length, the cumulative resistance heat of the welding wire under the current current-carrying state is calculated.

[0085] It should be noted that the preset resistance thermophysical model of the welding wire is based on Joule's law, and the input parameters of the model are determined: resistivity of the welding wire material, diameter and cross-sectional area of ​​the welding wire, real-time extension length, welding current and energizing time, and resistance is established. With heat production The fundamental mathematical relationships are established. Then, through welding experiments under different extension lengths and current conditions, the temperature change of the welding wire and the heat loss coefficient are measured to correct errors caused by actual working conditions such as environmental heat dissipation and droplet transfer. Finally, a quantitative model that can be embedded in the controller and can directly substitute real-time monitoring data to calculate the cumulative resistance heat is formed.

[0086] If the wire extension length is detected to be greater than the preset maximum safety limit, or the accumulated resistance heat is greater than the preset burnout warning threshold, the controller sends a compensation pulse sequence to the wire feeding drive mechanism to force the extension length to be restored, and drives the eccentric rotary motor set inside the handle to generate vibration feedback.

[0087] It should be noted that the maximum safety limit value is set by combining welding process tests and equipment operating boundaries. First, through actual measurements under different current and voltage, welding wire specifications and extension lengths, the upper limit of critical parameters for preventing the welding wire from burning out, the arc from extinguishing, and the contact tip from overheating is determined. Then, based on the allowable range of welding deviation, the safe range of heat input to the molten pool, and the safety margin reserved for the movement limits of the mechanism, the limit value is finally written into the controller as the upper limit for preventing over-adjustment and ensuring welding and equipment safety.

[0088] It should be noted that the burnout warning threshold is set by combining the welding wire material, diameter, welding current and real-time extension length. First, the heat value of the welding wire resistance heat accumulation to the melting critical value is determined by test under different working conditions. Then, factors such as heat dissipation conditions and arc fluctuations are added for correction, and a safety margin is reserved. Finally, the critical heat value is written into the controller as the warning threshold, which is used to compare the calculated resistance heat accumulation in real time to realize early warning and protection against welding wire burnout risk.

[0089] If the wire extension length is less than or equal to the preset minimum safety limit, the controller forcibly cuts off the movement command of the conductive tip axial drive mechanism and locks the conductive tip in the current position to prevent the wire from physically sticking to the conductive tip.

[0090] In a preferred embodiment of the present invention, the method further includes feedforward compensation logic, and the specific steps are as follows: the controller monitors the effective value of the welding current in real time, and calculates the theoretical melting rate of the welding wire under the current current heat input level based on the physical laws of metal melting and using a preset melting rate calculation rule.

[0091] It should be noted that the calculation method for welding wire melting speed is based on Joule's law and the thermal balance relationship. A quantitative calculation model is established by comprehensively considering parameters such as welding current, welding wire resistivity, extension length, cross-sectional area, and arc heat input. The core of the formula is that the melting speed is approximately proportional to the square of the current. Corrections are made by incorporating the latent heat of fusion of the welding wire material, the cumulative resistance heat, the wire feed speed, and the heat loss coefficient. By inputting the current operating parameters in real time, the instantaneous melting speed of the welding wire under current-carrying conditions can be obtained, providing a matching basis for the coordinated adjustment of wire feed and contact tip displacement.

[0092] When performing dynamic adjustment of the wire feeding speed, the controller adds the theoretical melting rate as a basic normal compensation term to the control command.

[0093] The sum of the adjustment component calculated based on the deviation vector and the feedforward compensation component obtained based on the theoretical melting rate is used as the final output control quantity to the wire feeding drive mechanism.

[0094] The above content is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the concept of the invention or exceed the scope defined by the present invention, and all such modifications or additions should fall within the protection scope of the present invention.

Claims

1. A semi-automatic manual welding high-efficiency track-finding auxiliary system, characterized in that: include: The environmental data acquisition module uses a miniature laser triangulation sensor to obtain the real-time physical distance between the end of the welding wire and the bottom of the bevel on the workpiece surface. The distance deviation calculation module compares the real-time physical distance with a preset reference distance benchmark and calculates the deviation vector value. The wire extension adjustment module, based on the deviation vector value, dynamically adjusts the wire extension length by changing the wire feeding speed and the physical displacement of the contact tip in the axial direction. The pointing correction control module utilizes the effect of the change in the directionality of the arc force caused by the change in the extension length of the welding wire to generate a lateral restoring force at the end of the welding wire pointing towards the center line of the bevel. The track-seeking status monitoring module collects infrared thermal distribution images of the molten pool and the rear area, obtains the centroid coordinates of the molten pool, compares them with the expected center position calculated through distance information, and determines the validity of the current track-seeking status. The control priority analysis module collects lateral pressure data applied by the welder through a pressure sensor array, determines the welder's operating intention, and adjusts the control priority of automatic track finding accordingly. The closed-loop safety protection module monitors the current and voltage parameters in the welding circuit in real time. When abnormal fluctuations or exceeding the safe range are detected, an alarm signal is issued and the drive mechanism is locked.

2. The semi-automatic manual welding high-efficiency track-finding auxiliary system according to claim 1, characterized in that: The specific method for obtaining the real-time physical distance is as follows: The total physical distance from the sensor reference plane to the workpiece surface is collected by using a miniature laser triangulation range sensor. The fixed physical deviation from the sensor reference plane to the end of the welding wire is subtracted to obtain the real-time physical distance value from the end of the welding wire to the bottom of the bevel. During the measurement process, the controller initiates high-frequency sampling logic to perform weighted averaging of distance values ​​within multiple consecutive sampling periods. The sampling points with time nodes closer to the current moment are assigned higher weight coefficients. The moving average algorithm filters out instantaneous interference jumps caused by arc fluctuations or metal droplet splashes during the welding process.

3. The semi-automatic manual welding high-efficiency track-finding auxiliary system according to claim 1, characterized in that: The specific calculation method for the deviation vector value is as follows: A reference distance benchmark database for plates of different thicknesses and different bevel types is pre-built and stored; Before the welding operation begins, the corresponding reference distance reference value is automatically matched in the reference distance reference database based on the input plate thickness, bevel angle and preset initial extension length of welding wire. Once the welding process begins, the controller extracts the filtered real-time physical distance at fixed time intervals and subtracts the reference distance baseline value from it to obtain the difference. If the difference is positive, it is determined that the physical position of the welding wire tip relative to the bottom of the groove has been raised, and it is identified as the trajectory of the manual welding gun deviating from the center line of the groove and shifting towards the side wall. If the difference is negative, it is determined that the end of the welding wire has sunk, which is identified as the end of the welding wire being too close to the surface of the workpiece. The controller defines the directionality of the deviation based on the positive or negative attribute of the difference, and defines the intensity attribute of the deviation based on the absolute value of the difference, thus synthesizing a deviation vector with both directionality and intensity attributes.

4. The semi-automatic manual welding high-efficiency track-finding auxiliary system according to claim 1, characterized in that: The specific method for dynamically adjusting the welding wire extension length is as follows: A collaborative conversion model between wire feeding speed and conductive tip displacement is established within the controller; When the deviation vector value indicates that the wire extension length needs to be increased, the controller outputs a first command to the wire feeding drive mechanism to reduce the pulse frequency of the wire feeding motor to slow down the rate at which the wire is fed into the molten pool. At the same time, it outputs a second command to the conductive tip axial drive mechanism to drive the micro stepper motor to retract the conductive tip in a direction away from the workpiece through a precision lead screw pair. When the deviation vector value indicates that the wire extension length needs to be reduced, the controller outputs a third command to increase the pulse frequency of the wire feeding motor to accelerate the wire replenishment, and at the same time outputs a fourth command to drive the micro stepper motor to extend the conductive tip forward toward the workpiece.

5. The semi-automatic manual welding high-efficiency track-finding auxiliary system according to claim 1, characterized in that: The specific logic for generating the lateral restoring force is as follows: Based on the positive correlation between the wire extension length and resistance heat in gas metal arc welding, when the wire extension length increases, the resistance heat component of the wire itself increases, thereby changing the plasma hydrodynamic distribution of the arc at the end of the wire. When the wire extension length exceeds the preset process stability value, the axial thrust component generated by the arc weakens, while the transverse force component acting on the bevel sidewall is enhanced due to the expansion of the plasma column, resulting in a physical tendency to push the wire end towards the bevel sidewall. When the wire extension length is less than the preset process stability value, the arc energy is concentrated in the center area of ​​the bottom of the bevel, generating an attractive restoring force pointing towards the root of the bevel. By finely adjusting the wire extension length using a controller, the spatial distribution change of the arc force is transformed into a restoring torque pointing towards the center line of the bevel.

6. The semi-automatic manual welding high-efficiency track-finding auxiliary system according to claim 1, characterized in that: The specific method for determining the validity of the current track-seeking state is as follows: The infrared imaging component eliminates visible light interference through a long-wave infrared filter and aligns the center of its field of view with the liquid molten pool area behind the welding arc. The controller acquires raw infrared data frames containing the temperature field distribution of the molten pool, performs grayscale processing on the raw infrared data frames, and uses a preset grayscale threshold to segment the image into regions, identifying continuous regions with grayscale values ​​higher than the high temperature threshold as the target areas of the molten pool. Calculate the arithmetic mean of the horizontal and vertical coordinates of all pixels within the target area of ​​the molten pool, and define it as the centroid coordinate of the molten pool. The controller extracts the projection position of the centroid coordinates of the molten pool onto the transverse axis and compares it in real time with the theoretical bevel center coordinates calculated by the miniature laser triangulation sensor through geometric transformation. If the spatial deviation between the two values ​​exceeds the preset safety deviation limit, the controller determines that the current laser ranging data is obstructed by oxides or impurities on the bevel surface, causing abnormal ranging. Subsequently, it outputs a command to forcibly pause the automatic adjustment logic and lock the current wire feeding parameters until the infrared imaging component confirms that the centroid coordinates of the molten pool have returned to the logic allowable range.

7. The semi-automatic manual welding high-efficiency track-finding auxiliary system according to claim 1, characterized in that: The specific method for determining the welder's operational intent is as follows: Multiple thin-film pressure gauges are embedded in the left and right gripping areas of the manual welding gun handle to form a pressure sensor array; The controller collects the analog voltage signals of each diaphragm pressure gauge in real time, and calculates the pressure difference between the total pressure on the left and the total pressure on the right by weighted summation, which is recorded as the lateral operating pressure deviation. The lateral operating pressure deviation signal is fed into a digital low-pass filter with a preset cutoff frequency for smoothing, eliminating high-frequency components generated by the physiological tremors of the welder's hands. The filtered pressure difference is compared with a preset intervention threshold. If the absolute value of the pressure difference is greater than the intervention threshold, it is determined that the welder is performing an active trajectory correction operation. The controller automatically reduces the gain coefficient of the automatic track-finding control loop to put the automatic adjustment system in a follow-up state.

8. The semi-automatic manual welding high-efficiency track-finding auxiliary system according to claim 5, characterized in that: The method also includes a pre-calibration process for the transfer function of distance and restoring force, the specific steps of which are as follows: For V-groove, U-groove and I-groove, establish independent experimental calibration databases in advance under controlled conditions; During the calibration process, by fixing the manual welding gun and manually changing the wire extension length, a three-dimensional force sensor is used to measure the numerical changes of the transverse arc force on the wire end under different bevel geometries, forming a multi-dimensional lookup table model. The lookup table model uses the bevel angle, bevel depth, real-time welding current, real-time welding voltage, and current wire extension length as multi-dimensional input variables, and the expected lateral restoring force as the output variable. During actual online welding, the controller calls the corresponding lookup table model in real time based on the identified bevel type, and calculates the precise adjustment amount of the wire extension length required under the current instantaneous working condition through interpolation algorithm.

9. The semi-automatic manual welding high-efficiency track-finding auxiliary system according to claim 1, characterized in that: The specific analysis process of the closed-loop security protection module is as follows: The electrical and physical parameters of the welding power supply output terminal are extracted in real time through the Hall current sensor and voltage sampling circuit. The product of the instantaneous value of welding current and the instantaneous value of welding voltage is calculated and recorded as the real-time energy input value of welding. Using a preset resistance thermophysical model of the welding wire, combined with the real-time monitored value of the welding wire extension length, the cumulative amount of resistance heat of the welding wire under the current current-carrying state is calculated. If the wire extension length is detected to be greater than the preset maximum safety limit, or the accumulated resistance heat is greater than the preset burnout warning threshold, the controller sends a compensation pulse sequence to the wire feeding drive mechanism to force the extension length to be restored, and drives the eccentric rotary motor set inside the handle to generate vibration feedback. If the wire extension length is less than or equal to the preset minimum safety limit, the controller forcibly cuts off the movement command of the conductive tip axial drive mechanism and locks the conductive tip in the current position to prevent the wire from physically sticking to the conductive tip.

10. The semi-automatic manual welding high-efficiency track-finding auxiliary system according to claim 9, characterized in that: The method also includes feedforward compensation logic, the specific steps of which are as follows: The controller monitors the effective value of the welding current in real time and, based on the physical laws of metal melting, calculates the theoretical melting rate of the welding wire at the current current heat input level using a preset melting rate calculation rule. When performing dynamic adjustment of the wire feeding speed, the controller adds the theoretical melting rate as a basic normal compensation term to the control command; The sum of the adjustment component calculated based on the deviation vector and the feedforward compensation component obtained based on the theoretical melting rate is used as the final output control quantity to the wire feeding drive mechanism.