Airport terminal metal roof welding seam detection system and method

Abstract

The invention discloses an airport terminal metal roof welding seam detection system and method, and the system comprises a roof data collection device and a robot carrying the roof data collection device, and the roof data collection device is used for obtaining the surface data of left and right welding seams of the robot. The robot carries out auxiliary positioning on the relative track position of the robot through a main control system and a TOF laser range finder, the main control system is formed by connecting four independent single-chip microcomputers in a mutual communication mode, according to surface data of left and right weld joints of the robot, the defects of the terminal building metal roof are inspected, and the defects comprise broken welding, weld penetration, dislocation welding and spot welding. By adopting the technical scheme of the invention, an airport terminal metal roof inspection task is realized.

Description

Technical Field

[0001] The present invention belongs to the technical field of weld quality inspection, and particularly relates to a weld inspection system and method for the metal roof of a terminal building. Background Art

[0002] With the rapid development of the modern civil aviation industry, the construction scale of large airport terminals is constantly expanding, and the requirements for building quality are becoming increasingly strict. Due to its beautiful appearance, durability, corrosion resistance and other characteristics, stainless steel roofs have been widely used in the field of large buildings. However, the weld quality of stainless steel roofs is directly related to the safety and service life of buildings.

[0003] Traditional weld quality inspection methods mostly rely on manual visual inspection. This method is not only inefficient but also easily affected by human factors, resulting in low inspection accuracy. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a weld inspection system and method for the metal roof of a terminal building.

[0005] To achieve the above object, the present invention adopts the following technical solutions:

[0006] A weld inspection system for the metal roof of a terminal building includes: a roof data acquisition device and a robot carrying the roof data acquisition device. The roof data acquisition device is used to obtain the surface data of the welds on the left and right sides of the robot. The robot, through a main control system connected by four independent single-chip microcomputers communicating with each other and a TOF laser rangefinder, performs auxiliary positioning of the relative track position of the robot according to the surface data of the welds on the left and right sides of the robot, and realizes the inspection of defects on the metal roof of the terminal building.

[0007] Preferably, the main control circuit is composed of four subsystems with STM32 series single-chip microcomputers as the core. Among them, the STM32F103RCT6 single-chip microcomputer is used to receive the signals of four laser TOF modules and communicate with the STM32ZET6_1 single-chip microcomputer system through the CAN bus. The STM32ZET6_1 receives the CAN communication signal of the STM32RCT6, conducts serial communication with the STM32ZET6_2, and simultaneously sends signals to the roof data acquisition device. The STM32ZET6_2 conducts serial communication with the STM32ZET6_1 and the STM32F407VET7, processes the calculation of laser TOF ranging data, the calculation of robot state control, and the request for arranging the robot operation mode. The STM32F407 conducts serial communication with the STM32ZET6_2, receives the requested speed signal and returns the actual speed signal, generates a PWM waveform through PID to control the dual-motor drive, and realizes differential adjustment of the robot.

[0008] Preferably, STM32F407VET7 implements tasks such as motor operation, receiving actual data transmitted back by the encoder, displaying speed information, receiving speed commands from STM32F407ZET6_2, and calculating and controlling the PWM wave of the motor in the system.

[0009] Preferably, STM32F103ZET6_2 controls tasks such as laser TOF data operation, robot attitude calculation, robot path planning, issuing robot motion commands, and issuing data acquisition commands in the system; STM32F103ZET6_2 classifies the robot motion state into a correction mode, a straight-line mode, and an obstacle-crossing mode.

[0010] Preferably, the roof data acquisition device uses a camera module driven by a K210 chip, is placed on the left and right sides of the robot by a rigid fixing method, communicates with the serial port of the robot through the UART pin in the K210 chip, receives weld track data, and records the weld image in the SD card in the K210 system.

[0011] Preferably, the defects include: lack of weld, burn-through, misaligned weld, and spot weld.

[0012] The present invention also provides a method for detecting welds on the metal roof of a terminal building, including:

[0013] Step S1, the roof data acquisition device obtains the surface data of the welds on the left and right sides of the robot; wherein, the data acquisition device uses a camera module driven by a K210 chip;

[0014] Step S2, the robot, through the main control system and the TOF laser rangefinder connected by mutual communication of four independent single-chip microcomputers, assists in positioning the relative track position of the robot according to the surface data of the welds on the left and right sides of the robot, and realizes the inspection defects of the metal roof of the terminal building, and the defects include: lack of weld, burn-through, misaligned weld, and spot weld.

[0015] Preferably, the main control circuit consists of four subsystems with STM32 series single-chip microcomputers as the core. Among them, the STM32F103RCT6 single-chip microcomputer is used to receive signals from four laser TOF modules and communicate with the STM32ZET6_1 single-chip microcomputer system via the CAN bus. The STM32ZET6_1 receives the CAN communication signal from the STM32RCT6, conducts serial communication with the STM32ZET6_2, and simultaneously sends signals to the roof data acquisition device. The STM32ZET6_2 conducts serial communication with the STM32ZET6_1 and the STM32F407VET7, and processes laser TOF ranging data calculation, robot status control calculation, and robot operation mode arrangement requests. The STM32F407 conducts serial communication with the STM32ZET6_2, receives the requested speed signal and returns the actual speed signal, generates PWM waveforms through PID to control the dual-motor drive, and realizes differential adjustment of the robot.

[0016] The present invention obtains the surface data of the left and right welds of the robot through the roof data acquisition device. The robot, through the main control system connected by four independent single-chip microcomputers communicating with each other and the TOF laser rangefinder, assists in positioning the relative track position of the robot according to the surface data of the left and right welds of the robot, and realizes the inspection of defects on the metal roof of the terminal building. The defects include: broken welding, penetrated welding, misaligned welding, and spot welding. By adopting the technical solution of the present invention, the inspection task of the metal roof of the terminal building is realized. Brief Description of the Drawings

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only the embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained according to the provided drawings without creative efforts.

[0018] Figure 1 It is a schematic diagram of the main control system of the metal roof weld system of the terminal building in the embodiment of the present invention;

[0019] Figure 2 It is a schematic diagram of the movement route of the robot chassis;

[0020] Figure 3 It is a schematic diagram of the movement state of the robot;

[0021] Figure 4 It is a data processing flow chart of the STM32F407VET7;

[0022] Figure 5 It is a data processing flow chart of the STM32F103ZET6_2;

[0023] Figure 6 Schematic diagram for robot body state recognition;

[0024] Figure 7 Schematic diagram of the process for collecting metal roof weld data;

[0025] Figure 8 Data processing flowchart of the roof data collection device. Specific implementation manners

[0026] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0027] To make the above objects, features, and advantages of the present invention more obvious and understandable, the present invention will be further described in detail below in conjunction with the accompanying drawings and specific implementation manners.

[0028] Embodiment 1:

[0029] The embodiment of the present invention provides a weld detection system for the metal roof of a terminal building. It is designed for a roof structure with an average weld height of 2.7 cm, a weld spacing of 40 cm, and a slope of 0 - 30° in the metal roof of the terminal building. There is a roof data collection device and a robot carrying the roof data collection device. The roof data collection device is used to obtain the surface data of the left and right welds of the robot. The robot, through a main control system connected by four independent single-chip microcomputers communicating with each other and a TOF laser rangefinder, based on the surface data of the left and right welds of the robot, assists in positioning the relative position of the robot on the track, and realizes the inspection of defects in the metal roof of the terminal building. The defects include: broken welds, burn-throughs, misaligned welds, and spot welds.

[0030] As an implementation manner of the embodiment of the present invention, the robot is a two-wheel drive tracked robot with dimensions of 330*255*100 (mm) and a combined aluminum alloy shell with a self-weight of 2.2 Kg. It steers by the differential speed of two motors and uses aviation shock-absorbing rubber materials and rivets to fix the four-way laser TOF ranging module. Under normal working conditions, the robot can travel in a straight line at a speed of 0 - 0.7 M / S and perform obstacle-crossing body movements between two upright lock-edge welds with a slope less than 40°, and can send control commands to the data collection device.

[0031] As Figure 1As shown in the figure, the main control circuit consists of four subsystems with STM32 series single-chip microcomputers as the core. Among them, the STM32F103RCT6 single-chip microcomputer is used to receive the signals of four laser TOF modules and communicate with the STM32ZET6_1 single-chip microcomputer system through the CAN bus; the STM32ZET6_1 receives the CAN communication signals of the STM32RCT6, conducts serial communication with the STM32ZET6_2, and sends signals to the roof data acquisition device at the same time to ensure the correct weld information is collected; the STM32ZET6_2 conducts serial communication with the STM32ZET6_1 and the STM32F407VET7, and processes the calculation of laser TOF ranging data, the calculation of robot state control, and the request for arranging the robot operation mode; the STM32F407 conducts serial communication with the STM32ZET6_2, receives the requested speed signal and returns the actual speed signal, generates a PWM waveform through PID to control the dual-motor drive, so as to achieve the purpose of differential adjustment of the robot.

[0032] The distance between every two stainless steel welds on the terminal building roof is 400 mm, and the average weld height is 27 mm. In order to detect each weld, the roof data acquisition device needs to collect the data of each weld from the side of the weld.

[0033] As Figure 2 , 3 shown in the figure, the robot travels in a straight line parallel to the left and right welds, and the roof data acquisition device is used to collect the surface data of the left and right welds of the robot during the travel. When reaching the end of the plate, the robot will automatically turn and move forward, so as to cross the weld, reach the next track, and adjust the position of the robot at the start of the second track to prepare for the second round of straight-line travel and data acquisition. Repeat until the route covers the entire metal roof.

[0034] As Figure 4 shown in the figure, the STM32F407VET7 realizes the tasks of motor operation, receiving the actual data returned by the encoder, displaying the speed information, receiving the speed command of the STM32F407ZET6_2, and calculating and controlling the PWM wave of the motor in the system.

[0035] In the STM32F407VET7, the system first obtains the actual motion data from the motor encoder. The encoder signal is input to the timer or external interrupt pin of the STM32F407VET7, and the STM32F407VET7 accurately captures the pulse signal of the encoder through the timer or capture interrupt. Through the number of pulses of the encoder signal, the STM32F407VET7 calculates the actual rotation speed of the motor. After filtering and data processing of this speed information by the STM32F407VET7, the rotation speed display of the motor is updated in real time. These processed data will be displayed on the OLED screen, and the real-time motor speed information can be viewed.

[0036] The STM32F407VET7 also receives speed commands from the STM32F407ZET6_2 through communication with it. The STM32F407VET7 analyzes the received target speed data, compares it with the actual speed of the current motor, and calculates the error between the two. Based on this error, the STM32F407VET7 adjusts the output PWM signal through the PWM (Pulse Width Modulation) algorithm to control the motor speed to approach the target speed. This process performs feedback adjustment in real time to ensure that the motor can maintain or change its speed according to the command.

[0037] According to the calculated PWM wave, the STM32F407VET7 generates corresponding PWM signals through the timer or external output module. These signals are used to control the motor drive module, thereby changing the speed or direction of the motor. The duty cycle of the PWM signal determines the magnitude of the current driving the motor, which in turn affects the motor speed. The entire process is a closed-loop control, that is, by continuously collecting encoder feedback signals, calculating speed errors, and adjusting the PWM waveform to control the operation of the motor in real time.

[0038] The STM32F407VET7 adjusts the action mode of the motor according to the path planning commands of the upstream system (STM32F103ZET6_2). When receiving path planning or motion commands, the STM32F407VET7 adjusts the motion state of the motor according to different command types (such as correction mode, straight-line mode, and obstacle-crossing mode). For example, in the correction mode, the STM32F407VET7 may need to adjust the motor speed and steering to correct the robot's posture; in the obstacle-crossing mode, the STM32F407VET7 adjusts the output power and speed of the motor to ensure that the robot can smoothly cross the obstacle. Through close cooperation with the STM32F103ZET6_2, the STM32F407VET7 not only controls the basic motion of the motor but also can accurately execute various motion commands according to the motion state of the robot and the requirements of path planning.

[0039] As Figure 5 shown, the STM32F103ZET6_2 controls tasks such as laser TOF data operation, robot posture calculation, robot path planning, robot motion command issuance, and data acquisition command issuance in the system. To plan the motion path, the STM32F103ZET6_2 divides the robot motion state into correction mode, straight-line mode, and obstacle-crossing mode.

[0040] The correction mode follows the following algorithm:

[0041] When the robot is going straight at the center of the track, the position of the robot relative to the track is as Figure 6As shown, the laser TOF measurement distances at the front left, front right, rear left, and rear right of the robot are Distance left_forward , Distance right_forward , Distance left_back , Distance right_back respectively. The distance between two laser TOF rangefinders on one side is N, the width of the robot is W, and the length is L. Therefore, the AVG relative orbit inclination angle θ can be calculated as follows:

[0042]

[0043] The distance D from the center of mass of the robot to the central axis of the orbit is as follows:

[0044]

[0045] Assume that the absolute value of the angle threshold in the algorithm is T θ , and the absolute value of the distance threshold is T D . The following state machine can be obtained:

[0046]

[0047] Do not adjust and enter the straight-line mode

[0048] In the correction mode, the linear velocity Vx of the robot is 0. When the angle needs to be adjusted, the robot is given an angular velocity of Vz = ±0.261 rad / s. The power adjustment power strategy satisfies:

[0049]

[0050] Among them, Pbase is the base power, τ is the current load torque of the motor, and k torquek is the torque compensation coefficient.

[0051] When the robot is corrected to an error less than the angle threshold with the orbit, the robot enters the straight-line mode.

[0052] The straight-line mode follows the following algorithm:

[0053] To prevent the robot from deviating from the center of the orbit, the robot will continuously detect the angle θ with the left and right orbits and the distance D from the central axis during driving, and the straight-line algorithm satisfies:

[0054] Assume that the absolute value of the angle threshold in the algorithm is T θ , and the absolute value of the distance threshold is T D :

[0055]

[0056] Do not correct the angular velocity and continue the straight-line mode

[0057] As a differential tracked robot, its linear velocity Vx, angular velocity Vr, wheelbase L, and gear radius are related to the speeds of the left and right wheels as follows:

[0058]

[0059] In the straight-line mode, the linear velocity of the robot Vx = 0.4 m / s. When the angle needs to be adjusted, the robot is given an angular velocity of Vz = ±0.131 rad / s. When the robot corrects the error with the track to be less than the angle threshold, the robot does not correct the angular velocity and maintains the straight-line mode.

[0060] In the straight-line mode, the power output is closely related to the speed requirement of the robot, and a feedback closed-loop control of power and speed needs to be established:

[0061] P = kv * (v target - v real )

[0062] Kv is the speed adjustment coefficient, v target is the target speed, and v real is the actual speed.

[0063] In the correction and straight-line modes of the robot, the above-mentioned mechanism for robot state recognition will be applied to ensure the normal operation of the robot between the tracks.

[0064] To ensure that the robot can turn at the end of the track, the robot uses an optical encoder as an odometer. The relationship between the optical encoder and the actual distance is that one grid of the encoder M = 0.002 m / grid. The single-chip microcomputer understands the travel of the robot by recording the encoder on the motor. Therefore, the turning condition is:

[0065]

[0066] Among them, N is the number of encoder grids recorded since the robot enters the straight-line mode, M is the corresponding relationship between the encoder and the actual distance, and D_roof is the length of the airport roof. The robot records its own mileage through the optical encoder and enters the obstacle-crossing mode when it reaches the roof mileage.

[0067] In the obstacle-crossing mode, the movement of the robot is divided into three stages:

[0068] The first stage: The linear velocity of the robot Vx = 0, and the angular velocity is Vz = ±1.57 rad / s. When the robot turns in place, the left and right tracks need to be driven in reverse, and the motor power should be set to medium to prevent excessive torque from causing the track to slip or damage the ground. The power adjustment power strategy in the first stage satisfies:

[0069] P = P base + k torque * |τ|

[0070] Among them, Pbase is the base power, τ is the current load torque of the motor, and k torquek is the torque compensation coefficient. After the first stage is executed for 2S, it enters the second stage;

[0071] Second stage: The linear velocity Vx of the robot is 0.2 m / s, and the angular velocity Vz is 0. When crossing an obstacle, higher power is required, and at the same time, the change in resistance needs to be monitored to prevent insufficient or excessive power.

[0072] When the robot crosses an obstacle, it needs to combine the inclination feedback of the IMU to dynamically adjust the power output.

[0073] Among them, the power needs to satisfy:

[0074] P = P base + k pitch *∣Pitch∣

[0075] When the inclination angle of the robot is large (for example, ∣Pitch∣> 10°), the power compensation kpitch*∣Pitch∣ is increased to ensure stability when crossing an obstacle. After the second stage is executed for 2S, it enters the third stage;

[0076] Third stage: After crossing the obstacle, the power needs to be immediately reduced to prevent rushing out of the track due to inertia. At the end of crossing the obstacle, the vehicle speed is controlled by braking. The dynamic power formula is as follows:

[0077] P target = P base - k brake *v

[0078] Among them, v is the current speed, and k brake is the braking coefficient.

[0079] At the moment when crossing the obstacle ends, the algorithm gradually reduces the power to a stable value by reducing the PWM duty cycle:

[0080] P = P targe t *(1 - e -λt )

[0081] Among them, λ is the adjustment coefficient, and t is the time after crossing the obstacle ends.

[0082] Subsequently, the linear velocity Vx of the robot is given as 0, and the angular velocity Vz is ±1.57 rad / s. After executing for 2S, the obstacle crossing mode is completed and the correction mode is entered; when the robot turns in place, the left and right tracks need to be driven in the reverse direction, and the motor power should be set to medium to prevent excessive torque from causing the tracks to spin idly or damage the ground. The power adjustment strategy in the third stage satisfies:

[0083] P = P base + k torque *|τ|

[0084] Among them, P base is the base power, τ is the current load torque of the motor, and k torquek is the torque compensation coefficient. After the third stage is executed for 2S, it enters the correction mode.

[0085] As an implementation manner of the embodiment of the present invention, the roof data acquisition device is a weld quality data acquisition device based on K210, which can stably and clearly capture the metal roof weld during the operation of the robot.

[0086] The roof data acquisition device adopts a camera module driven by a K210 chip, which has a resolution of 2 million pixels. It is placed on both sides of the robot in a rigid fixing manner and communicates with the serial port of the robot through the UART pin in the K210 chip, receives the weld track data, and records the weld image in the SD card in the K210 system, as Figure 7 shown.

[0087] In the normal working state of the roof data acquisition device, various track weld defects will be recorded in the SD card. Subsequently, technicians will extract the data from the SD card and use deep learning and object detection algorithms to batch process the data, extract defect features, and generate relevant detection reports.

[0088] Through the roof data acquisition device, defects such as discontinuous welding, incomplete penetration of the weld, misaligned welding, and spot welding during the welding process can be discovered and corrected in a timely manner, ensuring the long-term durability and safety of the metal roof.

[0089] As Figure 8 shown, after K210 is powered on, it waits for the robot to transmit a shooting command, extracts the track position and number information from the command, and records them in the SD card. The received command signals are shown in Table 1:

[0090] Table 1

[0091] Frame header Track number Start recording flag Check bit Frame tail 0xFF 1Bit 1Bit 1Bit OxEF

[0092] After receiving a command, K210 will first verify the checksum. If the checksum is correct, indicating that the command is valid, K210 will extract the track number and start recording flag from the command. If the start recording flag is 1, K210 will continue to process and start image acquisition. If the check fails, K210 will ignore the command and wait for a valid command to arrive again. Once the start recording flag is 1, K210 will start image acquisition according to the received track number. The system will obtain track images from the camera at a frame rate of 12 - 15 FPS and name the image files according to the track number and photo number. Each frame of the image will be named in the format of track number_photo number, such as "1_001", "1_002", etc., for subsequent data management and processing.

[0093] K210 will store these image data in the SD card and ensure the normal operation of the file system. During the acquisition process, K210 may periodically check the storage space. If the storage is close to full capacity, the system will trigger a warning or automatically stop recording to avoid data loss. The stored image data not only provides an important basis for subsequent weld quality assessment but also provides reliable support for the robot to work on complex terrains.

[0094] The data acquisition device of K210 can ensure stable driving during the operation of the robot. Especially in the area of the standing seam weld with a slope less than 40 degrees, the robot can maintain a stable speed of 0 to 0.7 meters per second. Through the robot's ability to cross the roof welds, stable operation in complex environments is ensured.

[0095] Overall, K210 not only efficiently acquires images while accumulating a large amount of weld quality data but also provides a scientific basis for subsequent weld detection, maintenance, and management, ultimately contributing to ensuring the long-term safe operation of large buildings such as the metal roof of the terminal building.

[0096] Embodiment 2:

[0097] The present invention also provides a method for detecting welds on the metal roof of a terminal building, including:

[0098] Step S1, the roof data acquisition device acquires the surface data of the left and right welds of the robot; wherein, the data acquisition device uses a camera module driven by a K210 chip;

[0099] Step S2, the robot uses a main control system and a TOF laser rangefinder connected by four independent single-chip microcomputers communicating with each other to assist in positioning the relative track position of the robot according to the surface data of the left and right welds of the robot, and realizes the inspection of defects on the metal roof of the terminal building. The defects include: broken welds, burn-throughs, misaligned welds, and spot welds.

[0100] As an implementation mode of the embodiment of the present invention, the main control circuit is composed of four subsystems with STM32 series single-chip microcomputers as the core; among them, the STM32F103RCT6 single-chip microcomputer is used to receive signals from four laser TOF modules and communicate with the STM32ZET6_1 single-chip microcomputer system through the CAN bus; the STM32ZET6_1 receives the CAN communication signal of the STM32RCT6, communicates with the STM32ZET6_2 through the serial port, and at the same time sends a signal to the roof data acquisition device; the STM32ZET6_2 communicates with the STM32ZET6_1 and the STM32F407VET7 through the serial port, processes the calculation of laser TOF ranging data, the calculation of robot state control, and the request for arranging the robot operation mode; the STM32F407 communicates with the STM32ZET6_2 through the serial port, receives the requested speed signal and returns the actual speed signal, generates a PWM waveform through PID to control the dual-motor drive, and realizes differential adjustment of the robot.

[0101] The above-described embodiments are only descriptions of the preferred embodiments of the present invention, and do not limit the scope of the present invention. Without departing from the design spirit of the present invention, various deformations and improvements made by those of ordinary skill in the art to the technical solution of the present invention shall fall within the protection scope determined by the claims of the present invention.

Claims

1. A terminal metal roof weld detection system, characterized in that: include: A roof data acquisition device and a robot equipped with the roof data acquisition device. The roof data acquisition device is used to obtain the surface data of the left and right welds of the robot. The robot uses a main control system composed of four independent single-chip microcomputers that communicate with each other and a TOF laser rangefinder to assist in positioning the robot relative to the track based on the surface data of the left and right welds of the robot, thereby realizing the inspection of defects on the metal roof of the terminal building.

2. The terminal metal roof weld detection system according to claim 1, characterized in that: The main control circuit is composed of four subsystems with STM32 series microcontrollers as the core; among them, the STM32F103RCT6 microcontroller is used to receive four-channel laser TOF module signals and communicate with the STM32ZET6_1 microcontroller system through the CAN bus; STM32ZET6_1 receives the CAN communication signal of STM32RCT6, and communicates with STM32ZET6_2 through serial port, and sends signals to the roof data acquisition device at the same time; STM32ZET6_2 communicates with STM32ZET6_1 and STM32F407VET7 through serial port to process laser TOF ranging data calculation, robot state control calculation, and robot operation mode arrangement request; STM32F407 communicates with STM32ZET6_2 through serial port, receives the request speed signal and returns the actual speed signal, generates PWM waveform through PID to control the dual motor drive, and realizes the differential adjustment robot.

3. The terminal metal roof weld detection system as claimed in claim 2, characterized in that: The STM32F407VET7 implements the tasks of running the motor, receiving actual data from the encoder, displaying speed information, receiving speed instructions from the STM32F407ZET6_2, and calculating the PWM wave to control the motor in the system.

4. The terminal metal roof weld detection system as claimed in claim 3, characterized in that: STM32F103ZET6_2 controls the tasks of laser TOF data calculation, robot posture calculation, robot path planning, robot motion command issuance and data acquisition command issuance in the system; STM32F103ZET6_2 divides the robot motion state into correction mode, straight mode and obstacle crossing mode.

5. The terminal metal roof weld detection system as claimed in claim 4, characterized in that: The roof data acquisition device uses a camera module driven by the K210 chip, which is rigidly fixed on the left and right sides of the robot. It communicates with the robot's serial port through the UART pin in the K210 chip, receives weld track data, and records the weld image in the SD card in the K210 system.

6. The terminal metal roof weld detection system as claimed in claim 5, characterized in that: The defects include: broken welds, weld penetration, misaligned welds and spot welds.

7. A method for detecting welds of metal roofs of terminal buildings, characterized in that: include: Step S1, the roof data acquisition device acquires the surface data of the left and right welds of the robot; wherein the data acquisition device adopts a camera module driven by a K210 chip; Step S2, the robot uses a main control system composed of four independent single-chip microcomputers connected to each other and a TOF laser rangefinder to assist in positioning the robot relative to the track according to the surface data of the left and right welds of the robot, so as to realize the inspection of defects of the metal roof of the terminal building, including broken welds, weld penetrations, misaligned welds and spot welds.

8. The terminal metal roof weld detection method according to claim 7, characterized in that: The main control circuit is composed of four subsystems with STM32 series microcontrollers as the core; among them, the STM32F103RCT6 microcontroller is used to receive four-channel laser TOF module signals and communicate with the STM32ZET6_1 microcontroller system through the CAN bus; STM32ZET6_1 receives the CAN communication signal of STM32RCT6, and communicates with STM32ZET6_2 through serial port, and sends signals to the roof data acquisition device at the same time; STM32ZET6_2 communicates with STM32ZET6_1 and STM32F407VET7 through serial port to process laser TOF ranging data calculation, robot state control calculation, and robot operation mode arrangement request; STM32F407 communicates with STM32ZET6_2 through serial port, receives the request speed signal and returns the actual speed signal, generates PWM waveform through PID to control the dual motor drive, and realizes the differential adjustment robot.

Images