A mobile intelligent welding robot for seat production

Abstract

The application relates to the technical field of robot welding, in particular to a mobile intelligent welding robot for seat production, which comprises an autonomous mobile chassis for autonomous navigation and obstacle avoidance in a seat production workshop, a multi-degree-of-freedom mechanical arm for realizing complex spatial trajectory welding, a welding executor for completing high-precision welding work, a visual identification system for capturing a welding seam position in real time and detecting welding quality and feeding back to a control system, and a central controller for coordinating operation of various modules; through the above steps, the robot is flexibly moved in the production workshop through the autonomous mobile chassis and SLAM navigation technology, so that the utilization rate of the equipment is improved and the floor area of the production line is reduced; in combination with the multi-degree-of-freedom mechanical arm and the high-precision visual identification system, the complex welding seam trajectory of a seat framework can be adaptively identified, and pose deviation in the welding process can be compensated in real time through force feedback, so that the consistency of welding seam forming quality is ensured.

Description

Technical Field

[0001] This invention relates to the field of robotic welding technology, and more particularly to a mobile intelligent welding robot for seat production. Background Technology

[0002] In the field of seat manufacturing, welding is a key process to ensure structural strength and assembly accuracy. Currently, the industry generally uses fixed welding workstations or handheld welding equipment to complete the welding of seat frames.

[0003] Existing fixed welding workstations lack flexibility and are difficult to adapt to the flexible production needs of multi-variety and small-batch production. They also occupy a large area, resulting in low workshop space utilization. Furthermore, operator welding suffers from unstable efficiency and weld quality that depends on the operator's experience. In particular, in the welding of complex curved surfaces (such as the connection between the headrest bracket and the backrest), defects such as incomplete welding or false welding are prone to occur, which affects product reliability. In addition, traditional welding equipment lacks real-time quality monitoring capabilities and often requires destructive sampling inspection after welding, which increases production costs and time delays.

[0004] Therefore, in response to the problems of insufficient flexibility and poor quality stability of the aforementioned welding equipment, this invention proposes a mobile intelligent welding robot for seat production. Through autonomous navigation, visual guidance, and adaptive welding technology, it realizes flexible layout of seat production line and closed-loop control of welding quality. Summary of the Invention

[0005] To overcome the problems of insufficient flexibility and poor quality stability of existing welding equipment, this invention proposes a mobile intelligent welding robot for seat production.

[0006] The technical solution of this invention is: a mobile intelligent welding robot for seat production, comprising:

[0007] The autonomous mobile chassis is used for autonomous navigation and obstacle avoidance within the seat production workshop, and supports seamless switching between multiple workstations;

[0008] Multi-degree-of-freedom robotic arms are used to achieve welding of complex spatial trajectories;

[0009] Welding actuators, integrated into the end effector of a robotic arm, are used to perform high-precision welding operations;

[0010] The visual recognition system is used to capture the weld position and detect the welding quality in real time, and then feed the feedback to the control system.

[0011] The central controller coordinates the operation of each module and optimizes the welding path and process parameters through algorithms.

[0012] The welding actuator is an adaptive welding torch, which integrates a current and voltage closed-loop control module. It automatically switches between pulse welding and continuous welding modes according to the weld type identified by the vision system. The welding torch head is equipped with a contact-type positioning probe, which corrects the three-dimensional coordinate deviation by lightly touching the workpiece surface before welding. The welding torch gas path has a built-in flow sensor to adjust the protective gas flow rate in real time.

[0013] The robot also includes a force feedback module, which contains a six-dimensional force sensor and an impedance control algorithm. It is installed between the end flange of the robotic arm and the welding torch to monitor the contact force and torque during the welding process in real time. When the workpiece assembly error or thermal deformation causes the welding torch to deviate from the preset trajectory, the robotic arm corrects its posture online based on the force feedback data. At the same time, it records the force over-limit event to the fault database for process optimization analysis.

[0014] Preferably, the autonomous mobile chassis adopts an omnidirectional wheel structure and has a built-in SLAM algorithm. It uses LiDAR and IMU sensors to build a real-time workshop environment map and combines a dynamic path planning module to avoid personnel and equipment, enabling the robot to accurately stop in narrow workstations. The chassis is equipped with an electromagnetic braking device to automatically lock the position during welding to eliminate displacement errors caused by the movement of the robotic arm. It also supports wireless communication with the AGV scheduling system and receives welding task sequences assigned by the central control console.

[0015] Preferably, the visual recognition system consists of a high frame rate industrial camera and a line laser scanner. The camera identifies the weld feature points of the seat frame based on a convolutional neural network and generates three-dimensional point cloud data. The scanner simultaneously detects the weld bevel angle and gap width. The central controller dynamically adjusts the welding torch travel speed and wire feed accordingly. After welding is completed, the system automatically marks the defect location and generates a quality report by comparing the molten pool image with the standard process library.

[0016] Preferably, the central controller connects to the factory's MES system via the OPC UA protocol, receives order data including seat model and weld number, calls the local welding parameter library to match current, voltage and welding wire type, and iteratively optimizes parameters based on historical welding quality data. The controller has an embedded real-time operating system.

[0017] Preferably, the robot also includes an audible and visual alarm device, which is integrated on the top of the robot and includes an RGB warning light and a buzzer. When the wire balance detector or gas pressure sensor triggers the threshold, the warning light switches to a red flashing mode and sends a warning signal to the central control console. At the same time, the specific fault code is displayed through the robotic arm HMI interface, and maintenance personnel can retrieve the emergency handling manual for the fault by scanning the code.

[0018] Preferably, the robot also includes a temperature sensor embedded inside the harmonic reducer of each joint of the robotic arm. The temperature sensor uses a PT100 thermocouple to collect temperature rise data in real time. When the temperature of any joint exceeds the safety threshold, the controller immediately reduces the movement speed and starts the air cooling system. If the temperature does not drop within 10 minutes, the robot is forced to stop and a maintenance work order is generated to prompt for checking the lubrication status or harmonic wear.

[0019] Preferably, the robot also includes a quick-change interface, which is an integrated pneumatic and electrical module. The welding gun head can be replaced in seconds through an electromagnetic locking mechanism. The interface has built-in self-cleaning contacts to prevent signal interference caused by oxidation. After the robot is replaced, it automatically calls the calibration program and uses a reference ball fixed on the tooling to calibrate the center point of the tool. The positioning error is controlled within ±0.05mm, which supports the mixed production line requirements of argon arc welding gun, spot welding clamp and laser welding head.

[0020] Preferably, the robot is interconnected with the MES system via a 5G private network. The MES dynamically assigns welding tasks to idle robots and monitors their working status. The system backend integrates a digital twin module to map robot position, welding quality, and energy consumption data in real time. Production managers can adjust production schedules through a visual dashboard, and the system automatically generates weld quality traceability reports for ISO auditing.

[0021] The beneficial effects of this invention are:

[0022] 1. By using an autonomous mobile chassis and SLAM navigation technology, the robot can move flexibly in the production workshop and quickly respond to the welding needs of different workstations, thereby significantly improving the utilization rate of the equipment and reducing the floor space of the production line. Combined with a multi-degree-of-freedom robotic arm and a high-precision vision recognition system, it can adaptively identify the complex weld trajectories of the seat frame and compensate for the positional deviations in the welding process in real time through force feedback, thereby ensuring the consistency of weld formation quality.

[0023] 2. The central controller dynamically optimizes process parameters based on the welding parameter library and real-time quality inspection data, thereby effectively avoiding the problem of operators' reliance on experience in welding, thus significantly reducing the defect rate of missing welds, incomplete welds and other defects, and thus achieving simultaneous improvement in welding efficiency and quality stability. Attached Figure Description

[0024] Figure 1 The diagram shown is a schematic representation of the system framework of the present invention.

[0025] Figure 2 The diagram shown illustrates the workflow of this invention. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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 some embodiments of the present invention, but 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.

[0027] Please see Figure 1 The present invention provides an embodiment: a mobile intelligent welding robot for seat production, comprising:

[0028] The autonomous mobile chassis is used for autonomous navigation and obstacle avoidance within the seat production workshop, and supports seamless switching between multiple workstations;

[0029] A multi-degree-of-freedom robotic arm is used to achieve welding of complex spatial trajectories and adapt to the curved surface structure of the seat frame;

[0030] Welding actuators, integrated into the end effector of a robotic arm, are used to perform high-precision welding operations such as arc welding and laser welding.

[0031] The visual recognition system is used to capture the weld position and detect the welding quality in real time, and then feed the feedback to the control system.

[0032] The central controller coordinates the operation of each module and optimizes the welding path and process parameters through algorithms.

[0033] The autonomous mobile chassis adopts an omnidirectional wheel structure and has a built-in SLAM algorithm. It uses LiDAR and IMU sensors to build a real-time map of the workshop environment and combines a dynamic path planning module to avoid obstacles such as personnel and equipment, enabling the robot to accurately stop in narrow workstations. The chassis is equipped with an electromagnetic braking device, which automatically locks the position during welding to eliminate displacement errors caused by the movement of the robotic arm. It also supports wireless communication with the AGV scheduling system and receives welding task sequences assigned by the central control console.

[0034] The welding actuator is an adaptive welding torch with an integrated current and voltage closed-loop control module. It automatically switches between pulse welding and continuous welding modes based on the weld type (such as flat weld, fillet weld, or lap weld) identified by the vision system. The welding torch head is equipped with a contact-type positioning probe, which corrects the three-dimensional coordinate deviation by lightly touching the workpiece surface before welding. The welding torch gas path has a built-in flow sensor to adjust the shielding gas flow rate in real time to match the welding requirements of different metal materials (such as low carbon steel or aluminum alloy).

[0035] The visual recognition system consists of a high frame rate industrial camera and a line laser scanner. The camera uses a convolutional neural network (CNN) to identify the weld feature points of the seat frame and generate three-dimensional point cloud data. The scanner simultaneously detects the weld bevel angle and gap width. The central controller dynamically adjusts the welding torch travel speed and wire feed accordingly. After welding is completed, the system automatically marks the location of defects such as porosity and undercut by comparing the molten pool image with the standard process library and generates a quality report.

[0036] The robot also includes a force feedback module, which contains a six-dimensional force sensor and an impedance control algorithm. It is installed between the end flange of the robotic arm and the welding torch to monitor the contact force and torque during the welding process in real time. When the workpiece assembly error or thermal deformation causes the welding torch to deviate from the preset trajectory, the robotic arm corrects its posture online based on the force feedback data. At the same time, it records the force over-limit event to the fault database for process optimization analysis.

[0037] The central controller connects to the factory's MES system via the OPC UA protocol, receives order data including seat model and weld number, calls the local welding parameter library to match current, voltage and welding wire type, and iteratively optimizes parameters based on historical welding quality data. The controller has an embedded real-time operating system (RTOS) to ensure the synchronous execution of robotic arm motion control, sensor data processing and communication tasks, with a delay error of less than 1ms.

[0038] The robot also includes an audible and visual alarm device, which is integrated on the top of the robot and includes an RGB warning light and a buzzer. When the wire balance detector or gas pressure sensor triggers the threshold, the warning light switches to a red flashing mode and sends a warning signal to the central control console. At the same time, the specific fault code (such as "E102 - Wire exhausted") is displayed through the robotic arm HMI interface. Maintenance personnel can retrieve the emergency handling manual for this fault by scanning the code.

[0039] The robot also includes a temperature sensor embedded inside the harmonic reducer of each joint of the robotic arm. The temperature sensor uses a PT100 thermocouple to collect temperature rise data in real time. When the temperature of any joint exceeds the safety threshold, the controller immediately reduces the movement speed and starts the air cooling system. If the temperature does not drop within 10 minutes, the robot is forced to stop and a maintenance work order is generated to prompt for checking the lubrication status or harmonic wear.

[0040] The robot also includes a quick-change interface, which is an integrated pneumatic and electrical module. It achieves the replacement of welding gun heads in seconds through an electromagnetic locking mechanism. The interface has built-in self-cleaning contacts to prevent signal interference caused by oxidation. After the robot is replaced, it automatically calls the calibration program and uses a reference ball fixed on the tooling to calibrate the tool center point (TCP). The positioning error is controlled within ±0.05mm, supporting the mixed production line requirements of argon arc welding guns, spot welding clamps and laser welding heads.

[0041] The robot is interconnected with the MES system via a 5G private network. The MES dynamically assigns welding tasks to idle robots and monitors their working status. The system backend integrates a digital twin module to map robot position, welding quality, and energy consumption data in real time. Production managers can adjust production schedules through a visual dashboard. The system automatically generates weld quality traceability reports for ISO auditing.

[0042] Please see Figure 2 Furthermore, the workflow of this invention will be described in detail below:

[0043] The robot receives welding task instructions from the MES system via a wireless communication module, thereby parsing the seat model, weld location, and process parameters. It also calls upon the workshop environment map constructed by the SLAM module and combines it with real-time obstacle detection data to generate the optimal movement path. The robot then autonomously navigates to the target workstation and confirms the workpiece identity via a QR code or RFID tag upon arrival, thus ensuring that the welding object is correct.

[0044] The visual recognition system is activated, and a high-frame-rate industrial camera captures multi-angle images of the seat frame. Based on a deep learning algorithm, the feature points of the weld are extracted and three-dimensional coordinates are generated. At the same time, the line laser scanner simultaneously measures the weld bevel angle and gap width. The force sensor at the end of the robotic arm lightly touches the workpiece surface for contact positioning. The coordinates of the welding start point are corrected by combining data from multiple sensors, and the positioning accuracy reaches ±0.1mm.

[0045] The central controller calls the preset process parameter library according to the weld type (continuous welding, spot welding, etc.) to match the current, voltage and wire feed speed. The welding torch moves along the planned trajectory under the drive of the robotic arm. At the same time, the force feedback module monitors the contact force in real time and dynamically adjusts the position of the robotic arm to compensate for thermal deformation. During the welding process, the gas flow sensor and the arc monitoring module work together to ensure that the protective gas coverage is sufficient and the arc stability meets the standard.

[0046] After welding is completed, the vision system immediately performs a high-definition scan of the weld, detects defects such as porosity and undercut through image processing algorithms, and compares the results with the process standards. If a non-conforming weld is found, its location is marked and an audible and visual alarm is triggered. At the same time, the defect data is uploaded to the MES system to generate a rework order. For qualified welds, the welding parameters and quality data are automatically recorded for process optimization analysis.

[0047] The robot releases the workpiece fixture and retracts its robotic arm. The autonomously moving chassis plans the path to the next workstation according to the MES task queue. During the movement, it continuously monitors the dynamic obstacles in the environment. After arriving at the new workstation, the quick-change interface automatically switches the welding gun head (such as changing from argon arc welding to spot welding clamp). After completing the tool calibration, it enters a new round of welding cycle, thereby realizing the mixed-line continuous production of multiple seat models.

[0048] During operation, the robot collects real-time operational data through a sensor network distributed across key components. This data includes parameters such as the joint temperature of the robotic arm (monitored by a PT100 thermocouple to monitor the temperature rise of the harmonic reducer), the remaining amount of welding wire (detected by a laser rangefinder to check the remaining amount on the welding wire spool), the protective gas pressure (monitored by a digital pressure sensor to monitor the gas cylinder pressure), and fluctuations in servo motor current. The system also has preset multi-level warning thresholds: when the joint temperature exceeds 65°C, a level one warning is triggered, and the controller automatically reduces the robotic arm's movement speed and activates the air-cooling system; if the temperature continues to rise to 75°C and does not drop within 10 minutes, the warning is upgraded. For Level 2 alarms, the current task is immediately stopped and the robot is locked. At the same time, an emergency work order is sent to the maintenance terminal via the audible and visual alarm device (three-color warning light flashing red + buzzer sounding) and wireless network. The work order is automatically linked to the robot's historical maintenance records and fault handling guidelines. For consumable shortage warnings (such as welding wire remaining less than 10% or gas pressure below 0.5MPa), the system pushes a replenishment reminder to the warehouse management system 4 hours in advance to avoid production interruption. All warning events are recorded with accurate timestamps, fault codes and handling status, forming structured logs stored on local SSDs and backed up in the cloud.

[0049] The robot is equipped with an edge computing module to perform real-time compression and feature extraction on the entire welding process data (including robotic arm motion trajectory coordinates, welding current and voltage waveforms, original visual inspection images, force feedback curves, etc.), and synchronizes it to the factory's central database via a 5G private network. Data archiving employs a layered storage strategy: high-frequency process parameters retain their original data for 7 days before being downsampled and stored, while key quality indicators (weld defect types, welding duration, etc.) are permanently saved and linked to specific seat product serial numbers. Archived data is protected by blockchain technology to ensure immutability and supports third-party quality audit traceability. The system also includes a built-in analysis engine. It can automatically generate equipment health assessment reports (such as wear trend prediction of harmonic reducers), welding qualification rate statistical dashboards and energy consumption analysis charts, and use machine learning models to uncover the implicit correlation between process parameters and quality defects (such as identifying the negative correlation between crack defects in a batch of aluminum alloy seats and protective gas flow rate), thereby driving process optimization. Maintenance personnel can access robot working status curves for any time period or receive spare parts replacement suggestions based on predictive maintenance algorithms through a mobile APP (such as "the left arm joint bearing is expected to have a remaining life of 120 hours"), thereby achieving an upgrade from passive maintenance to proactive maintenance.

[0050] Through the above steps, the robot can move flexibly in the production workshop by using an autonomous mobile chassis and SLAM navigation technology, which significantly improves the utilization rate of the equipment and reduces the floor space of the production line. Combined with a multi-degree-of-freedom robotic arm and a high-precision vision recognition system, it can adaptively identify the complex weld trajectories of the seat frame and compensate for the positional deviations in the welding process in real time through force feedback, thereby ensuring the consistency of weld formation quality and greatly reducing the defect rate of missing welds and incomplete welds. This achieves a simultaneous improvement in welding efficiency and quality stability, thus solving the problems of insufficient flexibility and poor quality stability of existing welding equipment.

Claims

1. A mobile intelligent welding robot for seat production, characterized in that, Including: The autonomous mobile chassis is used for autonomous navigation and obstacle avoidance within the seat production workshop, and supports seamless switching between multiple workstations; Multi-degree-of-freedom robotic arms are used to achieve welding of complex spatial trajectories; Welding actuators, integrated into the end effector of a robotic arm, are used to perform high-precision welding operations; The visual recognition system is used to capture the weld position and detect the welding quality in real time, and then feed the feedback to the control system. The central controller coordinates the operation of each module and optimizes the welding path and process parameters through algorithms. The welding actuator is an adaptive welding torch, which integrates a current and voltage closed-loop control module. It automatically switches between pulse welding and continuous welding modes according to the weld type identified by the vision system. The welding torch head is equipped with a contact-type positioning probe, which corrects the three-dimensional coordinate deviation by lightly touching the workpiece surface before welding. The welding torch gas path has a built-in flow sensor to adjust the protective gas flow rate in real time. The robot also includes a force feedback module, which contains a six-dimensional force sensor and an impedance control algorithm. It is installed between the end flange of the robotic arm and the welding torch to monitor the contact force and torque during the welding process in real time. When the workpiece assembly error or thermal deformation causes the welding torch to deviate from the preset trajectory, the robotic arm corrects its posture online based on the force feedback data. At the same time, it records the force over-limit event to the fault database for process optimization analysis.

2. The mobile intelligent welding robot for seat production according to claim 1, characterized in that: The autonomous mobile chassis adopts an omnidirectional wheel structure and has a built-in SLAM algorithm. It uses LiDAR and IMU sensors to build a real-time map of the workshop environment. Combined with a dynamic path planning module, it avoids personnel and equipment, enabling the robot to accurately stop in narrow workstations. The chassis is equipped with an electromagnetic braking device, which automatically locks the position during welding to eliminate displacement errors caused by the movement of the robotic arm. It also supports wireless communication with the AGV scheduling system and receives welding task sequences assigned by the central control console.

3. The mobile intelligent welding robot for seat production according to claim 1, characterized in that: The visual recognition system consists of a high frame rate industrial camera and a line laser scanner. The camera identifies weld feature points of the seat frame based on a convolutional neural network and generates three-dimensional point cloud data. The scanner simultaneously detects the weld bevel angle and gap width. The central controller dynamically adjusts the welding torch travel speed and wire feed accordingly. After welding is completed, the system automatically marks the defect location and generates a quality report by comparing the molten pool image with the standard process library.

4. A mobile intelligent welding robot for seat production according to claim 1, characterized in that: The central controller connects to the factory's MES system via the OPC UA protocol, receives order data including seat model and weld number, calls the local welding parameter library to match current, voltage and welding wire type, and iteratively optimizes parameters based on historical welding quality data. The controller has an embedded real-time operating system.

5. A mobile intelligent welding robot for seat production according to claim 1, characterized in that: The robot also includes an audible and visual alarm device, which is integrated on the top of the robot and includes an RGB warning light and a buzzer. When the wire balance detector or gas pressure sensor triggers the threshold, the warning light switches to a red flashing mode and sends a warning signal to the central control console. At the same time, the specific fault code is displayed through the robotic arm HMI interface, and maintenance personnel can retrieve the emergency handling manual for the fault by scanning the code.

6. A mobile intelligent welding robot for seat production according to claim 1, characterized in that: The robot also includes a temperature sensor embedded inside the harmonic reducer of each joint of the robotic arm. The temperature sensor uses a PT100 thermocouple to collect temperature rise data in real time. When the temperature of any joint exceeds the safety threshold, the controller immediately reduces the movement speed and starts the air cooling system. If the temperature does not drop within 10 minutes, the robot is forced to stop and a maintenance work order is generated to prompt for checking the lubrication status or harmonic wear.

7. A mobile intelligent welding robot for seat production according to claim 1, characterized in that: The robot also includes a quick-change interface, which is an integrated pneumatic and electrical module. It achieves the replacement of welding gun heads in seconds through an electromagnetic locking mechanism. The interface has built-in self-cleaning contacts to prevent signal interference caused by oxidation. After the robot is replaced, it automatically calls the calibration program and uses a reference ball fixed on the tooling to calibrate the center point of the tool. The positioning error is controlled within ±0.05mm, which supports the mixed production line requirements of argon arc welding guns, spot welding pliers and laser welding heads.

8. A mobile intelligent welding robot for seat production according to any one of claims 1-7, characterized in that: The robot is interconnected with the MES system via a 5G private network. The MES dynamically assigns welding tasks to idle robots and monitors their working status. The system backend integrates a digital twin module to map robot position, welding quality, and energy consumption data in real time. Production managers can adjust production schedules through a visual dashboard. The system automatically generates weld quality traceability reports for ISO auditing.

Images