Welding task execution method, device and equipment and storage medium

Abstract

The invention discloses a welding task execution method, device and equipment and a storage medium, relates to the technical field of data processing, is applied to a control center and comprises the steps that welding tasks are issued to all robots according to initial positioning information of the robots; welding task execution results returned after the robots execute the welding tasks and current positioning information of the robots are obtained; the welding task execution result comprises a welding seam image and initial position information of a welding point; a preset reinforcement learning model is used for conducting correction processing on the welding task execution result, and target position information is obtained; and on the basis of the current positioning result of each robot and the target position information corresponding to the corresponding welding task, the welding task is issued to each robot again so as to execute the welding task. Thus, the position information, fed back by the robot, of the welding point can be corrected through the reinforcement learning model, the welding continuity and accuracy can be guaranteed in combination with the precise welding position, and the welding efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a welding task execution method, device, equipment and storage medium. Background Art

[0002] For welding tasks in manufacturing environments (automobile manufacturing, aerospace, large steel structures, precision equipment assembly), the welding tasks are sent to the robot in advance, and the robot performs the welding and feedbacks the corresponding welding results, that is, the location of the welding point. In the process of determining the location of the welding point, a monocular camera is usually used to capture the weld image, and a direct method is used instead of a feature point method to establish a large-scale semi-dense map; then the feature points extracted from the image are matched for inter-frame estimation. This processing has certain defects, and changes in lighting conditions will significantly affect the performance of the monocular camera. In low-light or strong light environments, the image quality may deteriorate, resulting in the failure of feature point detection and matching.

[0003] It can be seen that how to accurately locate the position of the welding point is a problem to be solved in this field. Summary of the invention

[0004] In view of this, the purpose of the present invention is to provide a welding task execution method, device, equipment and storage medium, which can correct the position information of the welding point fed back by the robot through the reinforcement learning model, and combine the precise welding position to ensure the continuity and accuracy of welding and improve the welding efficiency. The specific scheme is as follows:

[0005] In a first aspect, the present application provides a welding task execution method, which is applied to a control center, comprising:

[0006] Issue welding tasks to each robot according to the robot's initial positioning information;

[0007] Acquire the welding task execution result returned by each robot after executing the corresponding welding task and the current positioning information of the robot; the welding task execution result includes the weld image and the initial position information of the corresponding welding point;

[0008] The welding task execution result is corrected by using a preset reinforcement learning model to obtain target position information of the corresponding welding point;

[0009] Based on the current positioning result of each robot and the target position information corresponding to the corresponding welding task, the welding task is re-issued to each robot so that each robot can perform the corresponding welding task.

[0010] Optionally, the weld image is a weld image containing a laser strip projected on a welding area by a laser and captured by a visual sensor after the robot completes a corresponding welding task.

[0011] Optionally, the initial position information is the initial position information of the corresponding welding point in the coordinate system of the visual sensor calculated by the robot based on a laser triangulation method and the weld image;

[0012] The process of calculating and obtaining the initial position information includes:

[0013] Preprocessing the weld image, extracting characteristic points of the laser strip from the preprocessed weld image, and determining the intersection position of the laser plane and the weld plane according to the characteristic points;

[0014] The initial position information of the intersection position in the three-dimensional coordinate system of the visual sensor is calculated through a pinhole camera model.

[0015] Optionally, the using of a preset reinforcement learning model to correct the welding task execution result to obtain target position information of a corresponding welding point includes:

[0016] Using the convolutional layer of the preset reinforcement learning model to extract features of the weld image in the welding task execution result, to obtain a corresponding feature vector;

[0017] Based on the feature vector, the initial position information is coordinate adjusted using the reinforcement learning network of the preset reinforcement learning model to obtain target position information of the corresponding welding point.

[0018] Optionally, the reissuing a welding task to each robot based on the current positioning result of each robot and the target position information corresponding to the corresponding welding task includes:

[0019] Re-planning the welding tasks to be performed corresponding to the robots according to the current positioning results of the robots and the target position information corresponding to the corresponding welding tasks;

[0020] Based on the fast Fourier transform algorithm, each welding task to be executed is sent to the corresponding robot respectively.

[0021] In a second aspect, the present application provides a welding task execution method, which is applied to a robot, comprising:

[0022] Obtain the welding task issued by the control center based on the initial positioning information of the robot;

[0023] Execute the welding task, obtain a corresponding welding task execution result, and send the welding task execution result and its own current positioning information to the control center, so that the control center uses a preset reinforcement learning model to correct the welding task execution result to obtain the target position information of the corresponding welding point; the welding task execution result includes a weld image and the initial position information of the corresponding welding point;

[0024] The control center obtains the welding task re-issued to the robot based on the current positioning result of the robot and the target position information corresponding to the corresponding welding task, and executes the corresponding welding task.

[0025] In a third aspect, the present application provides a welding task execution device, which is applied to a control center, comprising:

[0026] A first task issuing module is used to issue welding tasks to each robot according to the initial positioning information of the robot;

[0027] A data acquisition module, used to acquire the welding task execution result returned by each robot after executing the corresponding welding task and the current positioning information of the robot; the welding task execution result includes a weld image and initial position information of the corresponding welding point;

[0028] A correction module, used to correct the welding task execution result by using a preset reinforcement learning model to obtain the target position information of the corresponding welding point;

[0029] The second task issuing module is used to re-issue the welding task to each robot based on the current positioning result of each robot and the target position information corresponding to the corresponding welding task, so that each robot can perform the corresponding welding task.

[0030] In a fourth aspect, the present application provides a welding task execution device, applied to a robot, comprising:

[0031] A first task acquisition module is used to acquire a welding task issued by the control center based on the initial positioning information of the robot;

[0032] a data sending module, used to execute the welding task, obtain the corresponding welding task execution result, and send the welding task execution result and its own current positioning information to the control center, so that the control center can use a preset reinforcement learning model to correct the welding task execution result to obtain the target position information of the corresponding welding point; the welding task execution result includes a weld image and the initial position information of the corresponding welding point;

[0033] The second task acquisition module is used to obtain the welding task re-issued to the robot by the control center based on the current positioning result of the robot and the target position information corresponding to the corresponding welding task, and execute the corresponding welding task.

[0034] In a fifth aspect, the present application provides an electronic device, including:

[0035] Memory, used to store computer programs;

[0036] A processor is used to execute the computer program to implement the welding task execution method as described above.

[0037] In a sixth aspect, the present application provides a computer-readable storage medium for storing a computer program, which, when executed by a processor, implements the welding task execution method as described above.

[0038] It can be seen that in this application, the control center can send welding tasks to each robot according to the initial positioning information of the robot; then obtain the welding task execution results returned by each robot after executing the corresponding welding task and the current positioning information of the robot; the welding task execution results include the weld image and the initial position information of the corresponding welding point; then use the preset reinforcement learning model to correct the welding task execution results to obtain the target position information of the corresponding welding point; then based on the current positioning results of each robot and the target position information corresponding to the corresponding welding task, re-send the welding task to each robot so that each robot performs the corresponding welding task. In this way, the present application can simultaneously control multiple robots to perform welding tasks in a coordinated manner, and correct the position information of the welding points fed back by the robot through the reinforcement learning model to obtain a more accurate welding position; further, the welding tasks to be performed can be re-planned according to the welding tasks performed by each robot and the position information of the robot, and sent to each robot for execution; combined with the accurate welding position, the continuity and accuracy of welding can be guaranteed, and the welding efficiency can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0040] Figure 1 A flow chart of a welding task execution method disclosed in this application;

[0041] Figure 2This is a schematic diagram of the incident angles of three pairs of antennas disclosed in this application;

[0042] Figure 3 A timing diagram of a specific transmission phase disclosed in this application;

[0043] Figure 4 A flowchart of another specific welding task execution method disclosed in this application;

[0044] Figure 5 A flowchart of another specific welding task execution method disclosed in this application;

[0045] Figure 6 This is a schematic structural diagram of a welding task execution device disclosed in this application;

[0046] Figure 7 This is a schematic structural diagram of another welding task execution device disclosed in this application;

[0047] Figure 8 This is a structural diagram of an electronic device disclosed in this application. DETAILED DESCRIPTION

[0048] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0049] See also Figure 1 As shown, an embodiment of the present invention discloses a welding task execution method, which is applied to a control center and includes:

[0050] Step S11: Send welding tasks to each robot according to the initial positioning information of the robot.

[0051] In an embodiment of the present application, the control center can manage multiple robots at the same time, and divide the sub-welding tasks corresponding to each robot according to the initial positioning information of the multiple robots and the overall welding work to be completed, and send the welding tasks to the robots respectively, so that the sub-welding tasks can be performed simultaneously by multiple robots. In a specific embodiment, each robot can be equipped with an ultra-wideband sensor and a 4-antenna array to communicate with each other, and the RTA (Reconfigurable Transmission Array) array of UWB (Ultra Wide Band, ultra-wideband wireless carrier communication technology) can be used, and the arrival phase difference (Phase Difference of Arrival, PDoA, a positioning algorithm based on phase difference) between different antenna pairs can be compared to determine the positioning information of the robot.

[0052] Step S12, obtaining the welding task execution result returned by each robot after executing the corresponding welding task and the current positioning information of the robot; the welding task execution result includes the weld image and the initial position information of the corresponding welding point.

[0053] In this embodiment, when each robot performs a welding task, it can obtain the initial position information of the welding point corresponding to the welding task and the corresponding weld image; further, the robot can share the execution result of its own welding task and its own current positioning information to the control center. In a specific embodiment, the weld image is a weld image containing a laser strip captured by a visual sensor after the robot completes the corresponding welding task by projecting a laser strip in the welding area through a laser. It can be understood that the robot can locate the weld by projecting a laser strip and capture the weld image through a visual sensor to achieve preliminary positioning of the weld. Further, the initial position information is the initial position information of the corresponding welding point in the coordinate system of the visual sensor calculated by the robot based on the laser triangulation method and the weld image; wherein the process of calculating the initial position information includes: preprocessing the weld image, extracting the feature points of the laser strip from the preprocessed weld image, and determining the intersection position of the laser plane and the weld plane according to the feature points; and calculating the initial position information of the intersection position in the three-dimensional coordinate system of the visual sensor through a pinhole camera model. Specifically, the robot can obtain the depth information of the weld by laser triangulation, and the preliminary positioning of the weld can be achieved by combining the weld image captured by the visual sensor; specifically, the laser projects a laser strip in the welding area to form an intersection with the weld surface; it is understandable that the projection angle and position of the laser strip can be accurately calibrated to ensure the accuracy of the measurement. Accordingly, the visual sensor (such as an industrial camera) can capture the weld image containing the laser strip; since the laser strip presents a high-contrast feature line in the image, the corresponding feature points can be further extracted later. Furthermore, the captured weld image needs to be preprocessed, which may include Gaussian blur and threshold segmentation, which can remove noise and highlight the feature points of the laser strip. The preprocessed image facilitates more accurate extraction of the edges and feature points of the laser strip. The feature points of the laser strip are then extracted from the preprocessed weld image to determine the intersection position of the laser plane and the weld surface. These intersection positions provide the depth information of the weld in space, which can further determine the initial position of the welding point. The coordinates of the intersection point P in the camera coordinates can be calculated using the following equation in combination with the pinhole model.

[0054] ;

[0055] Among them, (x c ,y c , z c ) is the position of P in camera coordinates; f is the focal length; ρ w and ρ h are the thickness and height of the image (in pixels); u and v represent the coordinates of the target point on the image plane; c z and c yare the correction parameters when converting from pixel coordinates to the camera coordinate system; (u0, v0) are the coordinates of the pixel at the center of the image. , b, c and d are the parameters of the laser plane equation.

[0056] Step S13: using a preset reinforcement learning model to correct the welding task execution result to obtain target position information of the corresponding welding point.

[0057] In the embodiment of the present application, the preprocessed weld image and the initial position information of the corresponding welding point obtained by the robot when performing the welding task can be obtained through the above steps; the initial position information is three-dimensional coordinate information, which may have certain errors and drifts. At this time, the reinforcement learning model can be further used to correct the initial position information to obtain more accurate target position information of the welding point. In a specific embodiment, the correction processing of the welding task execution result using the preset reinforcement learning model to obtain the target position information of the corresponding welding point can include: using the convolution layer of the preset reinforcement learning model to extract features of the weld image in the welding task execution result to obtain the corresponding feature vector; based on the feature vector, using the reinforcement learning network of the preset reinforcement learning model to adjust the coordinates of the initial position information to obtain the target position information of the corresponding welding point. Specifically, the preset reinforcement learning model includes a convolution filter and a deep Q network (combining deep learning and reinforcement learning) corresponding to the corresponding reinforcement learning algorithm; the preprocessed weld image is feature extracted through the convolution layer to generate a corresponding high-dimensional feature vector; these features capture the key patterns and details in the image, which can provide rich information for subsequent decision-making. Further, the subsequent action decision is made based on the reinforcement learning algorithm. Based on the extracted feature vector, the Q value of each possible action (such as the direction and amplitude of the moving tracking frame) can be calculated through the deep Q network; then the action with the highest Q value can be selected according to the greedy strategy to decide how to adjust the tracking position to reduce the error between the actual weld feature point. Further, according to the decision of the deep Q network, the initial position can be adjusted, such as moving one pixel upward; then the corrected coordinates are used for the next step of tracking and error calculation to gradually approach the actual weld position, and finally obtain accurate weld position information, that is, the target position information. It can be understood that the corrected three-dimensional coordinates have higher accuracy and stability, which can ensure that the welding path is consistent with the actual weld. In addition, the action instructions output by the deep Q network can be used to guide the corresponding robot to adjust the position of the welding point tracking frame to further optimize the tracking accuracy.

[0058] Step S14: Based on the current positioning result of each robot and the target position information corresponding to the corresponding welding task, re-issue the welding task to each robot so that each robot performs the corresponding welding task.

[0059] In this embodiment, the position correction of the welding point can be achieved through the above steps. The control center can re-plan the welding task of each robot based on the current positioning result of each robot and the corrected target position information of the welding point corresponding to the corresponding welding task, so that each robot can continue to perform the related welding task in collaboration. In a specific embodiment, a time multiplexing mechanism can be adopted. By quickly switching time slots, a robot can share data with the control center and / or other robots in different time slots; combining a specific time window to achieve data transmission sharing can avoid data conflict and data loss.

[0060] In a specific embodiment, the re-issuing welding tasks to each robot based on the current positioning results of each robot and the target position information corresponding to the corresponding welding task may include: re-planning the welding tasks to be performed corresponding to each robot according to the current positioning results of each robot and the target position information corresponding to the corresponding welding task; and respectively sending each welding task to be performed to the corresponding robot based on the fast Fourier transform algorithm. Specifically, during the data transmission between the control center and each robot, the calculation speed can be improved by fast Fourier transform, and the formula is as follows:

[0061] ;

[0062] Where x[n] is the input signal. N is the signal length. j is the imaginary unit. is the rotation factor.

[0063] Correspondingly, in a multi-robot scenario, tasks can be distributed among multiple robots through parallel computing:

[0064] ;

[0065] Among them, x1, x2, …x m It is the input signal between different robots.

[0066] In a specific embodiment, the control center may include a communication module, a data processing module, a task scheduling module and a user interface; wherein the communication module can realize wireless communication with each robot to ensure efficient transmission and synchronization of data; the data processing module can use the optimized phase difference estimation algorithm to calculate the relative position of each robot in real time. The task scheduling module can dynamically allocate and adjust the welding tasks of each robot according to the real-time positioning information of the robot. The user interface is an interface for the operator to monitor the system status in real time and manually intervene. The main functions of the control center include real-time collection of the position information of each robot, coordination of phase difference measurement requests, and integration and analysis of positioning data. It can be understood that considering a group of robots Move freely in a given environment without obstacles. The three-dimensional position of Each robot is equipped with an ultra-wideband sensor and a 4-antenna array to communicate with each other. The goal of each robot is to use this UWB system to obtain the relative position of its neighboring robots with respect to its own body frame. Estimates ,and , where R 3 For three-dimensional space, Indicates the rotation of the robot from the global to the body frame for relative positioning perception, which can calibrate the UWB ranging data more accurately. j is the body coordinate system of robot j, R j is the rotation matrix of robot j. Combined with an estimation algorithm that allows robot 𝑗 to sense its relative position to robot 𝑖 when robot 𝑖 initiates a ranging request, the relative positioning error can be minimized. The robot uses a set of ultra-wideband sensors to determine the distance and direction of the received communication signal. The robot may include an antenna array for sending and receiving data, a basic platform consisting of four UWB modules, capable of calculating the arrival phase of the input signal, and connected to a processing subsystem that performs all calculations. Furthermore, in order to effectively process the phase difference of multiple antenna pairs, the phase difference is used to estimate the relative position. The phase difference estimation algorithm is as follows:

[0067] ;

[0068] Among them, A is the phase difference matrix, A + is a pseudo-inverse matrix containing phase difference information between multiple antenna pairs. b is the observed signal vector. In order to determine the direction of the signal source, the RTA array of UWB can be used and the arrival phase difference between different antenna pairs can be compared. This operation can be performed to obtain the positioning of the signal source every time the robot initiates a relative position detection request. The ToA (Time of Arrival) of the radio signal between two nodes is calculated using asynchronous two-way ranging between nodes, which is mature and known to provide centimeter-level accuracy. For PDoA measurement, ideally, the distance between all antenna pairs is set to half the carrier wavelength λ / 2 wavelength to avoid wrapping (the estimated phase difference exceeds π or −π, thereby wrapping to the opposite sign), and the carrier frequency between all UWB transceivers should be synchronized.

[0069] Assume that the arrival phases of all 4 antennas in the RTA array are:

[0070] ;

[0071] The phase difference of all six antenna pairs can be further calculated using the following formula:

[0072] ;

[0073] in, The n and o in the formula represent any two antennas of the four antennas 1, 2, 3, and 4. n and o are not equal, and there are 6 combinations, i.e., 6 antenna pairs. is the arrival phase of the nth antenna of robot i and robot j, is the arrival phase of the oth antenna of robots i and j, represents the phase difference between the nth and oth antennas of robot j, where o and n are different values ​​in the same range. The phase difference calculated for sources perpendicular to the antenna pair may not be zero. The deviation compensation term can be used to compensate for this. The bias compensation can be found by first calibrating each antenna pair by measuring the phase angles for all possible angle pairs without any bias. These are related to the phase angles they should be based on their geometry and the bias factor The output true angle is compared to minimize the least squares error. After that, the formula for calculating the incident angle of all 6 pairs of antennas is as follows:

[0074] ;

[0075] in, is the incident angle of the antenna pair n and o of robots i and j relative to the signal source. is the phase difference measured between the antenna pairs n and o of robots i and j. The incident angles of the three pairs of antennas are as follows: Figure 2 As shown, the relative position relationship between the three pairs of antennas and the signal source is demonstrated. By calculating the incident angles between different antenna pairs, the precise positioning of the signal source can be achieved.

[0076] Using the six PDoA measurements and the specific geometry of the antenna array, the unit vector pointing to the source is obtained by solving the redundant equations: ;

[0077] ;

[0078] in, is the x-component of the unit direction vector from the antenna pair of robot i and robot j to the signal source; is the y-component of the unit direction vector from the antenna pair of robot i and robot j to the signal source; is the z-component of the unit direction vector from the antenna pair of robot i and robot j to the signal source; n and o in the equation represent any two antennas of 1, 2, 3, and 4, and n and o are not equal; x, y, and z are coordinate axes. Considering the characteristics of the UWB module, the noise and amplitude There is a direct correlation between the amplitude and the incident angle. When the amplitude is low, the estimated error of the incident angle is small; however, when the incident angle exceeds 70°, the estimation error increases significantly. Therefore, for a specific set of received phase differences, the output antenna pairs with very large phase differences are removed from the above equation. Experimental results show that when the threshold is set to =165°, the effect is good.

[0079] Next, based on the remaining pairs with valid phase difference values ​​in the system of equations, the pseudo-inverse of A (phase difference matrix) can be used and the equation below can be approximately solved to obtain the normalized direction estimate:

[0080] ;

[0081] in, represents the normalized direction vector in the mth measurement. represents the unnormalized direction vector in the mth measurement. It represents the sum of the squares of the components on the three coordinate axes x, y, and z. It can be understood that the complete AoA estimation relative positioning process can be implemented by Algorithm 1. The input of Algorithm 1 is the phase reaching each antenna, and the output is the direction of the adjacent unit vector. With the help of the above formula, Algorithm 1 can calculate the phase difference through offset compensation. , calculate the incident angle for each antenna pair , the unit direction vector The original estimate of , the unit vector Normalized to length 1. Further, using the distance measurement obtained by TWR and the direction obtained by Algorithm 1 , the estimated relative positioning is .

[0082] In a specific embodiment, a specific information transmission protocol is used between the robot and the RTA; Figure 3As shown in the figure, the timing relationship between the information transmission phase, the ranging phase and the orientation strategy phase is shown, with the information phase shown in blue, the ranging phase shown in red, and the orientation phase shown in green. When the device is used as a transmitter, only one antenna A4 is used. All steps start with the initialization of the information request, and the subsequent data transmission is divided into three phases: information transmission, ranging, and orientation measurement. In addition, the control center (such as a server) manages the execution of the welding tasks of each robot and dynamically allocates new welding tasks to improve the coordination and efficiency of the overall system. Further, in the information transmission phase, the information to be transmitted will be correctly formatted into 802.15.4A data packets and sent. In a typical 802.15.4A transmission, the maximum payload is 127 bytes. Therefore, the original information can be split into data packets of 120 bytes each for transmission, and the additional 7 bytes are used for cyclic redundancy check (CRC). This phase ensures that the necessary operation data, such as task status, position coordinates, and sensor data, can be exchanged efficiently and securely between the robots. In the ranging phase, the two-way ranging (TWR) protocol mentioned above can be used to achieve distance measurement to the source. In this phase, the receiver only uses its A4 receiver. Accurate distance measurement helps the relative positioning of the robots, ensuring accurate collision avoidance and optimization of the working path when working together. In the orientation measurement phase, that is, the phase of measuring the orientation between the transmitter and the receiver, all four antennas on the receiver side will be enabled. The orientation measurement data is used to further accurately position and enhance the collaboration accuracy between robots.

[0083] In another specific embodiment, in order to achieve efficient collaboration between multiple robots and ensure that each robot can receive new welding tasks in time and execute them accurately, the control center is used to implement the robot's welding task distribution and data integration. The control center is responsible for dynamically allocating new welding tasks according to production needs and the current status of each robot (such as task completion, position, load, etc.). By optimizing task allocation, the control center can improve the overall welding efficiency and shorten the production cycle. According to the urgency and priority of the task, the control center can reasonably arrange the execution order of the tasks to ensure that key tasks are completed first. The control center can monitor the working status of each robot in real time, including position, task progress, welding quality, etc., to ensure the overall coordination of the system. Exception handling: When a robot fails or the task execution is abnormal, the control center can detect and take corresponding emergency measures in time, such as reallocating tasks or notifying maintenance personnel. The control center can integrate sensor data and task execution information from all robots to form a comprehensive system status view. Performance optimization: By analyzing the collected data, the control center can identify system bottlenecks, optimize welding paths and task allocation strategies, and improve overall system performance. Correspondingly, during the information exchange process between the control center and the robot, after starting or completing the current task, the robot sends an initialization information request to the control center, reports its own status and requests a new welding task; the control center assigns new welding tasks according to the status and task requirements of each robot, and sends the task information to the corresponding robot through the 802.15.4A protocol; after receiving the task, the robot starts to perform the welding operation, and exchanges data with other robots and the control center during the information transmission, ranging and orientation measurement stages to ensure the smooth progress of the task; after completing the task, the robot sends the task completion status information to the control center, including completion time, welding quality assessment, etc. After receiving the feedback, the control center updates the task status and assigns new tasks or adjusts the existing task plan as needed; if an abnormality occurs during the execution of the task, the robot can report the fault information to the control center. The control center reallocates tasks or takes other emergency measures according to the situation to ensure that the production process is not affected.

[0084] In another specific embodiment, the relative positioning process of robot A is specifically as follows: Start relative position detection request: After completing the current welding task, robot A sends a relative position detection request to the control center to apply for relative position detection. Control center coordinates measurement: After receiving the relative position detection request, the control center sends a synchronous measurement signal to other robots (such as robots B, C, and D) to ensure that the antenna pairs of all relevant robots perform phase difference measurements at the same time. Phase difference measurement and data collection: Under the instruction of the control center, robot A and other robots use the RTA array of UWB to perform PDoA measurement and record the phase difference of each antenna pair. Data transmission and processing: All measurement data are uploaded through the control center, and the control center uses the optimized phase difference formula to calculate the relative position of robot A. Position update and feedback: After the calculation is completed, the control center updates the position information of robot A to the system database (of the control center), and feeds back the relevant position information to robot A and other robots for subsequent coordinated operations.

[0085] It can be seen that the present application can simultaneously control multiple robots to collaboratively perform welding tasks, and through the reinforcement learning model, the position information of the welding point fed back by the robot can be corrected to obtain a more accurate welding position, combined with deep reinforcement learning to improve welding accuracy, and use visual sensors and deep learning technology to track the weld in real time to ensure high-precision operation during welding. This can reduce the need for manual monitoring and shorten the production cycle; by integrating high-precision positioning sensors and real-time visual feedback systems, the relative position between the robot and the workpiece can be monitored in real time during the welding process, and the path and action of the robot can be dynamically adjusted; the time multiplexing mechanism is adopted to allow multiple robots to efficiently share positioning and welding information in a limited space, so that the robot can synchronously process information transmission and task allocation, optimize the overall welding process, and ensure the efficient cooperation of multiple robots in complex environments. Further, relying only on the robot's own sensors to achieve collaboration greatly reduces costs and complexity, and adapts to various complex construction environments; this can also ensure the accuracy of welding, and avoid positioning errors and physical deviations in traditional methods, greatly improve the intelligence level of the production line, reduce operating costs, and enhance production efficiency. Furthermore, the present application can re-plan the welding tasks that need to be performed based on the welding tasks completed by each robot and the position information of the robot, and send them to each robot for execution; combined with the precise welding position, the continuity and accuracy of the welding can be guaranteed, which can improve the welding efficiency.

[0086] like Figure 4 As shown, the embodiment of the present application discloses a welding task execution method, which is applied to a robot, comprising:

[0087] Step S21, obtaining the welding task issued by the control center based on the initial positioning information of the robot.

[0088] In the embodiment of the present application, the robot initially processes the waiting state. After receiving the initialization message sent by the control center, the robot can perform state initialization, that is, prepare to receive the welding task; the control center sends the initial welding task to each robot based on the initial positioning information of several robots and the welding tasks to be performed. It should be pointed out that when different robots perform welding tasks, they can feed back the task execution status and their own position information to other robots in order to coordinate the execution of welding tasks; during the process, when two robots transmit data, the control center can also be responsible for coordinating this communication process to ensure that only this pair of robots communicate within a specific time window to comply with the two-way communication restrictions of the UWB sensor.

[0089] Step S22, execute the welding task, obtain the corresponding welding task execution result, and send the welding task execution result and its own current positioning information to the control center, so that the control center uses a preset reinforcement learning model to correct the welding task execution result to obtain the target position information of the corresponding welding point; the welding task execution result includes a weld image and the initial position information of the corresponding welding point.

[0090] In this embodiment, after the robot performs the welding task and obtains the corresponding execution result, it can send the corresponding execution result and the current positioning information of the robot itself to the control center, and the control center will synchronize and coordinate the relevant information of other robots to ensure the stability of the working environment and the consistency of data; and the control center can correct the position information of the welding point based on the reinforcement learning model to obtain more accurate target position information. The process can refer to the content of the above embodiment and will not be repeated here.

[0091] Step S23, obtaining the welding task re-issued to the robot by the control center based on the current positioning result of the robot and the target position information corresponding to the corresponding welding task, and executing the corresponding welding task.

[0092] Furthermore, after correcting the position information of the welding point, the control center can re-divide the welding tasks that need to be performed based on the completion status of each welding task, and re-assign the welding tasks to each robot to ensure the welding quality and ultimately complete the welding task.

[0093] In a specific embodiment, Figure 5 As shown in the figure, in order to realize the collaborative execution of welding tasks by multiple robots, it is first necessary to configure the hardware of the robots to facilitate the three-dimensional positioning of the robots and the positioning of the welds when performing subsequent welding tasks. The corresponding weld positioning and tracking algorithm is combined to realize the execution of welding tasks and feed back relevant data to the control center. The control center manages multiple robots to realize multi-robot collaborative welding.

[0094] It can be seen that in the present application, multiple robots can collaboratively perform welding tasks, and the position information of the welding points fed back by the robots can be corrected through the reinforcement learning model of the control center to obtain a more accurate welding position; further, the control center can re-plan the welding tasks to be performed based on the welding tasks completed by each robot and the position information of the robot, and send them to each robot for execution; combined with the precise welding position, the continuity and accuracy of the welding can be guaranteed, which can improve the welding efficiency.

[0095] The following embodiment will specifically introduce the training process of the reinforcement learning model for correcting the initial position information of the welding point, which specifically includes:

[0096] In order to improve the accuracy of the weld position, deep reinforcement learning can be used to optimize the weld tracking process; specifically, the weld position can be obtained through the convolution filter, and the tracking position can be corrected using the deep Q network to ensure welding accuracy. This step continuously improves the weld tracking capability of the welding robot through the reinforcement learning model. It is understandable that due to the inevitable drift of the convolution-based filter tracker, there may be errors between the actual welding feature points and the position of the convolution filter tracker, which directly affects the quality of the welding process. In order to improve the accuracy of the weld feature points, reinforcement learning technology can be used here to improve the tracking position accuracy provided by the convolution filter tracker (CFT).

[0097] In order to train the deep reinforcement learning model, a virtual environment needs to be established. First, data preparation is performed. The data comes from the video recorded by the visual sensor when the robot moves and the welding gun is turned on, and the feature points of the weld are extracted. Next, a convolutional filter tracker is implemented to track the laser line features. Finally, the captured images (i.e., video frames) and the tracker position information are stored in a database. After the data preparation is completed, the data processing steps, including image preprocessing, labeling, and data augmentation, are performed to prepare the dataset for the training process of the agent. In the specific scenario, the image can be cropped to an appropriate size of 200×200 pixels and the coefficient bilinear interpolation method is used to resize the image to 100×100 pixels. Secondly, the collected data is labeled using VGG Image Annotator (VIA). In addition, data augmentation is performed to enrich the environment input and prevent overfitting when training the agent. The training dataset is enhanced by random horizontal flipping and random rotation. In addition, considering the dangerous situations such as weld bursting that may occur during welding, random brightness contrast and Gaussian noise techniques are used to improve the robustness and performance of the model. Therefore, a library of welding images and manually annotated ground truth feature points was constructed from more than 14,000 images containing position information. Subsequently, the dataset was divided into a training set and a test set, where the training data included 10,000 images and the test set consisted of the remaining images. It is worth noting that all images were from butt welds and involved different welding voltages and currents.

[0098] After completing the data set and successfully creating the virtual environment and deep Q learning model, the robot needs to be trained using the action-reward mechanism and gradient descent optimization method to maximize the reward function and minimize the loss. In this specific welding task application, the reward function can be expressed as:

[0099] ;

[0100] Among them, d1 and d2 represent the linear distance between the tracking point and the actual feature point before and after each action, is the reward function. After each step, the return of each operation is calculated as ,in, is the discount factor, and T-t+1 represents the remaining number of steps from the current time t to the termination time T. Then, the parameter update rule is based on the gradient descent method:

[0101] ;

[0102] in, and are the old and new neural network weights respectively; represents the learning rate, which is a hyperparameter that controls the step size when updating parameters; is the probability of performing the action, is its gradient, and They represent the state at time t and the action to be performed, respectively. Represents the neural network weights of the i-1th round of training. After each round, the neural network weights are updated . Parameter changes and action rewards G t , and the product of the probability gradient of the selected action, which is subdivided into the probability of performing the action. t The sign and magnitude of determine the direction and magnitude of parameter changes. The update rule in the following formula can be mathematically converted to optimize the following loss function:

[0103] ;

[0104] The mathematical expression of the update rule is as follows:

[0105] ;

[0106] in, is the learning rate, is the loss function Relative to the parameters The gradient of (neural network weight); r is the reward obtained by the robot after each step of the welding process; is a discount factor that weighs the impact of current rewards and future rewards. is the Q function of the old network, s and Indicates the current state and the current action to be performed; is the Q function of the new network representation, and Represent the next state and the possible action for the next step respectively. The deep learning model learned the policy network parameters through the above dataset. The state is input randomly and the batch size is 32. Each round has a maximum of 10 steps, or ends when the tracking error is less than 3 pixels (about 0.45 mm of the actual coordinates in the field of view of the visual system). After completing a round, a reward is returned to optimize the weights of the robot policy network. Each training cycle contains 1000 rounds. After 16,000 cycles, the reward gradually converges and the round length decreases. The final trained model can be used to correct the position of the weld image fed back by the robot and the position information of the corresponding welding point to obtain the corrected target position information.

[0107] It can be seen that this application combines laser triangulation and deep reinforcement learning technology, and the control center can accurately locate the weld position so as to adjust the welding trajectory in time to ensure the continuity and accuracy of welding. This automated weld tracking greatly reduces the need for manual monitoring, and multiple robots work in parallel, speeding up welding and shortening the production cycle.

[0108] like Figure 6 As shown, the embodiment of the present application discloses a welding task execution device, which is applied to a control center, including:

[0109] The first task issuing module 11 is used to issue welding tasks to each robot according to the initial positioning information of the robot;

[0110] A data acquisition module 12 is used to acquire the welding task execution result returned by each robot after executing the corresponding welding task and the current positioning information of the robot; the welding task execution result includes a weld image and initial position information of the corresponding welding point;

[0111] A correction module 13, used to correct the welding task execution result by using a preset reinforcement learning model to obtain target position information of the corresponding welding point;

[0112] The second task issuing module 14 is used to re-issue the welding task to each robot based on the current positioning result of each robot and the target position information corresponding to the corresponding welding task, so that each robot performs the corresponding welding task.

[0113] It can be seen that the present application can simultaneously control multiple robots to collaboratively perform welding tasks, and through the reinforcement learning model, the position information of the welding points fed back by the robots can be corrected to obtain a more accurate welding position; further, the welding tasks to be performed can be re-planned according to the welding tasks completed by each robot and the position information of the robots, and sent to each robot for execution; combined with the precise welding position, the continuity and accuracy of welding can be guaranteed, which can improve the welding efficiency.

[0114] In a specific embodiment, the correction module 13 may include:

[0115] A feature extraction unit, used to extract features of the weld image in the welding task execution result by using the convolution layer of the preset reinforcement learning model to obtain a corresponding feature vector;

[0116] The correction unit is used to adjust the coordinates of the initial position information based on the feature vector by using the reinforcement learning network of the preset reinforcement learning model to obtain the target position information of the corresponding welding point.

[0117] In a specific embodiment, the second task issuing module 14 may include:

[0118] A task planning unit, used for re-planning the welding tasks to be performed corresponding to each robot according to the current positioning results of each robot and the target position information corresponding to the corresponding welding task;

[0119] The task issuing unit is used to issue each of the welding tasks to be executed to the corresponding robots based on the fast Fourier transform algorithm.

[0120] like Figure 7 As shown, the embodiment of the present application discloses a welding task execution device, which is applied to a robot, including:

[0121] A first task acquisition module 21 is used to acquire a welding task issued by the control center based on the initial positioning information of the robot;

[0122] A data sending module 22 is used to execute the welding task, obtain the corresponding welding task execution result, and send the welding task execution result and its own current positioning information to the control center, so that the control center uses a preset reinforcement learning model to correct the welding task execution result to obtain the target position information of the corresponding welding point; the welding task execution result includes a weld image and the initial position information of the corresponding welding point;

[0123] The second task acquisition module 23 is used to obtain the welding task re-issued to the robot by the control center based on the current positioning result of the robot and the target position information corresponding to the corresponding welding task, and execute the corresponding welding task.

[0124] It can be seen that in the present application, multiple robots can collaboratively perform welding tasks, and the position information of the welding points fed back by the robots can be corrected through the reinforcement learning model of the control center to obtain a more accurate welding position; further, the control center can re-plan the welding tasks to be performed based on the welding tasks completed by each robot and the position information of the robot, and send them to each robot for execution; combined with the precise welding position, the continuity and accuracy of the welding can be guaranteed, which can improve the welding efficiency.

[0125] Furthermore, the present application also discloses an electronic device. Figure 8 This is a structural diagram of an electronic device 30 according to an exemplary embodiment. The content in the diagram cannot be regarded as any limitation on the scope of use of the present application.

[0126] Figure 8The present invention provides a schematic diagram of the structure of an electronic device 30 provided in an embodiment of the present application. The electronic device 30 may specifically include: at least one processor 31, at least one memory 32, a power supply 33, a communication interface 34, an input / output interface 35, and a communication bus 36. The memory 32 is used to store a computer program, which is loaded and executed by the processor 21 to implement the relevant steps in the welding task execution method disclosed in any of the aforementioned embodiments. In addition, the electronic device 30 in this embodiment may specifically be an electronic computer.

[0127] In this embodiment, the power supply 33 is used to provide working voltage for each hardware device on the electronic device 30; the communication interface 34 can create a data transmission channel between the electronic device 30 and the external device, and the communication protocol it follows is any communication protocol that can be applied to the technical solution of the present application, and is not specifically limited here; the input and output interface 35 is used to obtain external input data or output data to the outside world, and its specific interface type can be selected according to specific application needs and is not specifically limited here.

[0128] In addition, the memory 32, as a carrier for storing resources, can be a read-only memory, a random access memory, a disk or an optical disk, etc. The resources stored thereon can include an operating system 321, a computer program 322, etc., and the storage method can be temporary storage or permanent storage.

[0129] The operating system 321 is used to manage and control the hardware devices and computer programs 322 on the electronic device 30, which can be Windows Server, Netware, Unix, Linux, etc. In addition to the computer program that can be used to complete the welding task execution method performed by the electronic device 30 disclosed in any of the aforementioned embodiments, the computer program 322 can further include computer programs that can be used to complete other specific tasks.

[0130] Furthermore, the present application also discloses a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, the aforementioned disclosed welding task execution method is implemented. The specific steps of the method can refer to the corresponding contents disclosed in the aforementioned embodiments, and will not be repeated here.

[0131] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0132] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0133] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0134] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0135] The technical solution provided by the present application is introduced in detail above. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea. At the same time, for general technicians in this field, according to the idea of ​​the present application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A welding task execution method, characterized in that: Applied to control center, including: Issue welding tasks to each robot according to the robot's initial positioning information; Acquire the welding task execution result returned by each robot after executing the corresponding welding task and the current positioning information of the robot; the welding task execution result includes the weld image and the initial position information of the corresponding welding point; The welding task execution result is corrected by using a preset reinforcement learning model to obtain target position information of the corresponding welding point; Based on the current positioning result of each robot and the target position information corresponding to the corresponding welding task, the welding task is re-issued to each robot so that each robot can perform the corresponding welding task.

2. The welding task execution method according to claim 1, characterized in that: The weld image is a weld image containing a laser strip projected on the welding area by a laser and captured by a visual sensor after the robot completes the corresponding welding task.

3. The welding task execution method according to claim 2, characterized in that: The initial position information is the initial position information of the corresponding welding point in the coordinate system of the visual sensor calculated by the robot based on the laser triangulation method and the weld image; The process of calculating and obtaining the initial position information includes: Preprocessing the weld image, extracting characteristic points of the laser strip from the preprocessed weld image, and determining the intersection position of the laser plane and the weld plane according to the characteristic points; The initial position information of the intersection position in the three-dimensional coordinate system of the visual sensor is calculated through a pinhole camera model.

4. The welding task execution method according to claim 1, characterized in that: The method of using a preset reinforcement learning model to correct the welding task execution result to obtain target position information of the corresponding welding point includes: Using the convolutional layer of the preset reinforcement learning model to extract features of the weld image in the welding task execution result, to obtain a corresponding feature vector; Based on the feature vector, the initial position information is coordinate adjusted using the reinforcement learning network of the preset reinforcement learning model to obtain target position information of the corresponding welding point.

5. The welding task execution method according to claim 1, characterized in that: The re-issuing welding tasks to each robot based on the current positioning result of each robot and the target position information corresponding to the corresponding welding task includes: Re-planning the welding tasks to be performed corresponding to the robots according to the current positioning results of the robots and the target position information corresponding to the corresponding welding tasks; Based on the fast Fourier transform algorithm, each welding task to be executed is sent to the corresponding robot respectively.

6. A welding task execution method, characterized in that: Applications in robots include: Obtain the welding task issued by the control center based on the initial positioning information of the robot; Execute the welding task, obtain a corresponding welding task execution result, and send the welding task execution result and its own current positioning information to the control center, so that the control center uses a preset reinforcement learning model to correct the welding task execution result to obtain the target position information of the corresponding welding point; the welding task execution result includes a weld image and the initial position information of the corresponding welding point; The control center obtains the welding task re-issued to the robot based on the current positioning result of the robot and the target position information corresponding to the corresponding welding task, and executes the corresponding welding task.

7. A welding task execution device, characterized in that: Applied to control center, including: A first task issuing module is used to issue welding tasks to each robot according to the initial positioning information of the robot; A data acquisition module, used to acquire the welding task execution result returned by each robot after executing the corresponding welding task and the current positioning information of the robot; the welding task execution result includes a weld image and initial position information of the corresponding welding point; A correction module, used to correct the welding task execution result by using a preset reinforcement learning model to obtain the target position information of the corresponding welding point; The second task issuing module is used to re-issue the welding task to each robot based on the current positioning result of each robot and the target position information corresponding to the corresponding welding task, so that each robot can perform the corresponding welding task.

8. A welding task execution device, characterized in that: Applications in robots include: A first task acquisition module is used to acquire a welding task issued by the control center based on the initial positioning information of the robot; a data sending module, used to execute the welding task, obtain the corresponding welding task execution result, and send the welding task execution result and its own current positioning information to the control center, so that the control center can use a preset reinforcement learning model to correct the welding task execution result to obtain the target position information of the corresponding welding point; the welding task execution result includes a weld image and the initial position information of the corresponding welding point; The second task acquisition module is used to obtain the welding task re-issued to the robot by the control center based on the current positioning result of the robot and the target position information corresponding to the corresponding welding task, and execute the corresponding welding task.

9. An electronic device, characterized in that: include: Memory, used to store computer programs; A processor, configured to execute the computer program to implement the welding task execution method according to any one of claims 1 to 6.

10. A computer-readable storage medium, characterized in that: Used to store a computer program, which, when executed by a processor, implements the welding task execution method as described in any one of claims 1 to 6.

Images