A multi-process mechanical arm identification correction method and device

Abstract

This application relates to the field of industrial automation control technology, specifically a method and device for correcting deviations in multi-process robotic arm identification. By recording the current breakpoint pose and generating target instructions in real time when a high-priority emergency event is detected, preemptive intervention for high-risk anomalies is achieved. Simultaneously, based on the stored breakpoint pose, the robotic arm can be controlled to reset without damage and continue executing its intrinsic tasks after the emergency event is resolved. This avoids the missed detection of high-risk defective products and solves the problems of workflow disruption, logic deadlock, and missed actions caused by traditional hard interruptions. When the robotic arm performs a large-span temporary trajectory change in response to an emergency event, if the expected motion parameters exceed physical limits, the end effector posture is actively adjusted to replan a compliant trajectory, avoiding servo limit alarms, overload shutdowns, or collisions caused by forced trajectory changes. This improves the continuous operation rate of the robotic arm and equipment safety in complex concurrent environments.

Description

Technical Field

[0001] This application relates to the field of industrial automation control technology, specifically to a method and apparatus for correcting deviations in multi-process robotic arm identification. Background Technology

[0002] With the rapid development of industrial automation, the use of single or multiple robotic arms with vision recognition capabilities for global material handling and anomaly intervention has become the industry norm in complex, multi-process production line systems involving feeding, processing, discharging, and sorting. However, due to the asynchronous nature of vision recognition and the suddenness of non-standard abnormal events, existing robotic arm control systems have the following technical shortcomings when dealing with complex concurrent operating conditions: First, existing vision recognition and robotic arm control systems typically employ a rigid first-in-first-out (FIFO) queue to handle identified abnormal tasks. When the robotic arm is performing a lower-priority routine task or handling a general anomaly, such as a minor dimensional deviation, if the vision system suddenly identifies a very high-priority emergency, such as a high-risk defective product or foreign object intrusion, the system cannot dynamically preempt the interruption. This not only easily leads to missed detection of high-risk events or missed opportunities for optimal intervention, but also, if a forced interruption is executed directly, the system often cannot record the spatial state before the interruption, resulting in an inability to accurately resume the original task after the anomaly is handled, thus disrupting the original workflow.

[0003] Secondly, when the production line system forces the robotic arm to abandon its current preset path and temporarily plan a completely new grasping trajectory pointing to the unexpected event, the joint deflection angle of the robotic arm calculated by inverse kinematics is very likely to exceed its physical limit constraints or fall into a kinematic dead angle due to the sudden change in the target pose. Existing control methods usually directly report an error and stop the machine when the limit is exceeded, requiring manual intervention to reset, which reduces the automation and continuous operation rate of the production line.

[0004] Third, the robotic arm incurs additional time overhead during the handling of sudden interruptions, abnormal grasping, and trajectory reset, causing the actual execution cycle of the robotic arm to lag significantly behind the preset steady-state cycle. Existing control systems are mostly limited to the adjustment of individual robotic arms and cannot overcome the limits of single-machine physical acceleration. When the cycle lag is too large, since the external feeding and discharging conveyors still operate at their original fixed frequencies, the upstream and downstream production lines instantly lose cycle matching, which can easily lead to large-scale material accumulation, crushing, and impact, ultimately causing the collapse of the overall production rhythm.

[0005] Therefore, finding a suitable correction method for multi-process robotic arm recognition is a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0006] To address the aforementioned issues, this application provides a correction method and device for multi-process robotic arm identification, which can avoid missing high-risk defective products and solve the problems of workflow disruption, logic deadlock, and action omission caused by traditional hard interruption.

[0007] In a first aspect, this application provides a correction method for multi-process robotic arm identification, applied to a production line system. The production line system includes at least a robotic arm, an external feeding device, an external discharging device, and a controller. The method is executed by the controller and includes the following steps: Acquire multi-source data to construct a global state matrix, and control the robotic arm to execute a first-priority intrinsic task along a preset path based on the global state matrix; During the execution of the intrinsic task, a sudden event is identified based on the global state matrix. If the sudden event is evaluated as having a second priority and higher than the first priority, the current breakpoint pose is recorded and a target instruction for handling the sudden event is generated. When planning the motion trajectory according to the target instruction, if the expected motion parameters exceed the physical limits of the robotic arm, the end-effector posture in the target instruction is adjusted and the motion trajectory is replanned to drive the robotic arm to handle the sudden event with the new motion trajectory. After the emergency is handled, the robotic arm is controlled to continue executing the intrinsic task based on the breakpoint pose. Assess the cycle deviation rate caused by handling the sudden event, and trigger motion parameter compensation of the robotic arm or coordinated frequency reduction of the external feeding device and / or external discharging device based on the cycle deviation rate.

[0008] Furthermore, the multi-source data includes production line visual images, sensor detection signals, robotic arm encoder readings, and system clock data, and the global state matrix is ​​a data set containing three-dimensional coordinates of the production line space, material attribute status, and timestamp information.

[0009] Furthermore, the step of acquiring multi-source data to construct a global state matrix includes: Based on the system clock data, timestamps are uniformly assigned to the real-time acquired production line visual images, sensor detection signals, and encoder readings of the robotic arm body; The visual images of the production line with the same timestamp and the sensor detection signals are analyzed to determine the material attribute status. Combined with the preset equipment installation calibration parameters, the material attribute status is mapped to the three-dimensional spatial coordinates of the material in the production line coordinate system. The encoder readings of the robotic arm body with the same timestamp are obtained, and spatial pose mapping is performed in combination with the physical link parameters of the robotic arm to output the real-time three-dimensional spatial coordinates of the robotic arm end effector. The global state matrix is ​​generated by combining the material attribute status, the spatial three-dimensional coordinates of the material, and the real-time spatial three-dimensional coordinates of the robotic arm end effector at the same timestamp.

[0010] Furthermore, the step of controlling the robotic arm to perform a first-priority intrinsic task along a preset path based on the global state matrix includes: Extract the three-dimensional spatial coordinates of the target material from the global state matrix as the target grasping pose; The preset path is generated by combining the real-time three-dimensional spatial coordinates of the robotic arm's end effector to connect the current position with the target grasping posture. A drive command is sent to the robotic arm servo controller to drive the robotic arm to move along the preset path toward the target grasping pose.

[0011] Furthermore, the step of identifying sudden events based on the global state matrix includes: The material attribute states in the global state matrix are compared with a preset standard feature set in real time. If the material properties are found to meet the characteristics of high-risk defective products or foreign object intrusion, an abnormal signal is triggered. An abnormal state containing the characteristics of high-risk defective products or the characteristics of foreign object intrusion is identified as the sudden event.

[0012] Furthermore, the step of recording the current breakpoint pose includes: extracting the real-time physical deflection angle parameters of each joint of the robotic arm at the current moment, and the real-time three-dimensional spatial coordinates and three-dimensional attitude angles of the robotic arm end effector in the production line base coordinate system, and storing and protecting them as the breakpoint pose; The step of generating the target instruction for handling the emergency includes: extracting the real-time spatial three-dimensional coordinates and entity contour normal vector of the object corresponding to the emergency in the global state matrix; calculating the intervention angle without physical interference based on the entity contour normal vector as the desired end attitude target; and combining the intervention angle with the spatial target coordinates to generate the target instruction.

[0013] Furthermore, the step of planning the motion trajectory according to the target instruction includes: A kinematic mapping model is constructed based on the physical link parameters of the robotic arm; The end-effector posture target and spatial target coordinates in the target command are input into the kinematic mapping model for inverse pose mapping, and the expected physical deflection angle sequence of each joint is output to fit and generate the motion trajectory.

[0014] Furthermore, the step of adjusting the end effector posture in the target command and replanning the motion trajectory if the expected motion parameters exceed the physical limits of the robotic arm, so as to drive the robotic arm to handle the sudden event with the new motion trajectory, includes: When any joint angle in the expected physical deflection angle sequence exceeds the maximum mechanical deflection limit of that joint, the three-dimensional attitude angle corresponding to the expected end attitude target in the target command is compensated and adjusted. The adjusted end-effector pose target is re-input into the kinematic mapping model for inverse pose mapping until all joint angles are within the mechanical deflection limit, generating the compliant new motion trajectory. The new motion trajectory joint control command sequence is sent to the robotic arm servo controller.

[0015] Furthermore, the step of controlling the robotic arm to continue executing the intrinsic task based on the breakpoint pose includes: Extract the stored breakpoint pose; Plan a smooth obstacle avoidance backtracking trajectory from the current pose at the end of the handling of the emergency to the breakpoint pose; After driving the robotic arm to reset to the breakpoint pose along the smooth obstacle avoidance retreat trajectory, the remaining execution instructions of the intrinsic task before the termination are retrieved and executed along the preset path.

[0016] Furthermore, the step of assessing the cycle deviation rate caused by handling the sudden event, and triggering motion parameter compensation of the robotic arm based on the cycle deviation rate, or the coordinated frequency reduction of the external feeding device and / or external discharging device, includes: The additional time overhead caused by task interruption, emergency handling and reset process is statistically analyzed, and the cycle deviation rate of the current production line cycle time lagging behind the preset steady-state cycle time is calculated. When the cycle deviation rate is less than or equal to a preset coordination threshold, the joint drive speed of the robotic arm is increased in subsequent work cycles to perform the motion parameter compensation. When the cycle deviation rate is greater than the coordination threshold, a speed reduction control signal is sent to the external feeding device and / or external discharging device to synchronously reduce the running step cycle of the external material conveying mechanism until the cycle deviation rate returns to the preset error range.

[0017] Further, the collaborative threshold includes a first threshold and a second threshold, the first threshold being 5% and the second threshold being 10%, and the step of triggering motion parameter compensation of the robotic arm based on the cycle deviation rate, or the collaborative frequency reduction of the external feeding device and / or the external discharging device, includes: When the cycle deviation rate is less than or equal to 5%, the joint drive speed of the robotic arm is only finely adjusted in the subsequent work cycle, and the motion inflection points of the subsequent preset path are reduced. When the cycle deviation rate is greater than 5% and less than or equal to 10%, the joint drive speed of the robotic arm is continuously adjusted in the next two work cycles. When the cycle deviation rate is greater than 10%, the coordinated frequency reduction of the external feeding equipment and / or the external discharging equipment is triggered. The steps of the coordinated frequency reduction include: calculating the absolute delay time based on the cycle deviation rate, issuing a frequency reduction command containing the absolute delay time to the programmable logic controller of the external feeding device and / or external discharging device, proportionally reducing the operating frequency of the drive motor of the external material conveyor belt, or extending the start-stop step interval of the external material conveyor belt, so that the material input or output rate within the preset cycle matches the current actual processing cycle of the robotic arm.

[0018] Secondly, this application provides a correction device for multi-process robotic arm recognition, used for applying the correction method for multi-process robotic arm recognition as described in any one of the first aspects, the device comprising: The matrix construction and execution module is used to acquire multi-source data to construct a global state matrix, and control the robotic arm to execute a first-priority intrinsic task along a preset path based on the global state matrix; The breakpoint protection module is used to identify sudden events based on the global state matrix during the execution of the intrinsic task. If the sudden event is evaluated as having a second priority and higher than the first priority, the current breakpoint pose is recorded and a target instruction for handling the sudden event is generated. The attitude reconstruction and trajectory change module is used to adjust the end attitude in the target instruction and replan the motion trajectory when the expected motion parameters exceed the physical limits of the robotic arm, so as to drive the robotic arm to handle the sudden event with the new motion trajectory when the motion trajectory is planned according to the target instruction. The breakpoint resume module is used to control the robotic arm to continue executing the intrinsic task based on the breakpoint pose after the emergency event has been handled. The cycle time compensation module is used to evaluate the cycle time deviation rate caused by handling the sudden event, and trigger the motion parameter compensation of the robotic arm or the coordinated frequency reduction of the external feeding device and / or the external discharging device based on the cycle time deviation rate.

[0019] The method and apparatus for correcting deviations in multi-process robotic arm identification provided in this application have the following advantages: By recording the current breakpoint pose and generating target instructions in real time when a high-priority emergency event is detected, preemptive intervention for high-risk anomalies is achieved. Simultaneously, based on the stored breakpoint pose, the robotic arm can be controlled to reset without damage and continue executing its intrinsic tasks after the emergency event is handled. This avoids the missed detection of high-risk defective products and solves the problems of workflow disruption, logic deadlock, and missed actions caused by traditional hard interruptions. When the robotic arm performs a large-span temporary trajectory change in response to an emergency event, if the expected motion parameters exceed physical limits, the end effector posture is actively adjusted to replan a compliant trajectory, avoiding servo limit alarms, overload shutdowns, or collisions caused by forced trajectory changes, thus improving the continuous operation rate and equipment safety of the robotic arm in complex concurrent environments. In addition, by assessing and handling the cycle deviation rate caused by emergencies, it can not only trigger the mechanical arm's own motion parameter compensation, but also trigger the frequency reduction of external feeding and discharging equipment when the delay is severe, thus solving the problem of chain reactions such as severe accumulation and compression of upstream and downstream materials caused by abnormal time consumption of a single device. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a schematic diagram of the production line system in one embodiment; Figure 2 This is a flowchart illustrating a correction method for multi-process robotic arm identification in one embodiment; Figure 3 This is a schematic diagram of the main logic flow of a correction method for multi-process robotic arm identification in one embodiment; Figure 4 This is a flowchart illustrating the logic of spatial dimension trajectory correction in one embodiment. Figure 5 This is a schematic diagram of a multi-level clock compensation and cross-device collaborative frequency reduction process in one embodiment; Figure 6 This is a structural block diagram of a correction device for multi-process robotic arm identification in one embodiment; Figure 7 This is a structural block diagram of a processing device in one embodiment. Attached image description: 11. External feeding equipment; 12. External discharging equipment; 21. Robotic arm; 31. Production line vision sensor; 32. Photoelectric sensor; 33. Servo encoder; 34. Real-time operating system; 40. Electrical control box. Detailed Implementation

[0023] To enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0024] Example 1 In automated production lines with multiple processes and high concurrency, due to the asynchronous nature of visual inspection and the suddenness of non-standard anomalies, traditional control logic is prone to causing a chain of problems such as timing logic confusion, spatial track change shutdown, and global cycle time collapse when faced with sudden high-priority tasks interrupting the queue.

[0025] See Figure 1 , Figure 2 , Figure 3 , Figure 4 and Figure 5 As shown, to solve the above problems, this embodiment provides a correction method for multi-process robotic arm 21 identification, applied to a production line system. The production line system includes at least a robotic arm 21, an external feeding device 11, an external discharging device 12, and a controller. The method is executed by the controller and includes the following steps: S1. Acquire multi-source data to construct a global state matrix, and control the robotic arm 21 to perform a first-priority intrinsic task along a preset path based on the global state matrix; In step S1, by continuously acquiring and integrating multi-source data from the production line environment, a global state matrix reflecting the current real-time state is constructed. Based on the global state matrix, the robotic arm 21 is issued and monitored to perform routine intrinsic tasks.

[0026] S2. During the execution of the intrinsic task, a sudden event is identified based on the global state matrix. If the sudden event is evaluated as having a second priority and higher than the first priority, the current breakpoint pose is recorded and a target instruction for handling the sudden event is generated. In step S2, the global state matrix is ​​scanned in real time in the background. When a sudden event is detected and its priority is higher than the currently executing task, a logical interruption is implemented. At the same time, the spatial physical state of the robotic arm 21 at the moment of interruption, i.e., the breakpoint pose, is captured and locked, and new action instructions are generated for the sudden event. This ensures that high-priority events can be preemptively responded to immediately, while the original task's state is saved through breakpoint protection, preventing task state loss or control logic confusion.

[0027] S3. When planning the motion trajectory according to the target instruction, if the expected motion parameters exceed the physical limits of the robotic arm 21, the end posture in the target instruction is adjusted and the motion trajectory is replanned to drive the robotic arm 21 to handle the sudden event with the new motion trajectory. In step S3, if the expected joint angle exceeds the limit, the system actively relaxes non-critical attitude constraints and fine-tunes the end effector attitude angle. While ensuring accurate spatial positioning, it refits a new trajectory that conforms to the deflection limit of the robotic arm 21. This solves the problem of manual intervention required when the robotic arm 21 is prone to triggering servo alarms and being forced to stop due to kinematic dead angles during sudden, large-scale trajectory changes. By actively changing the minute attitude precision to obtain the spatial trajectory, it ensures stable operation of the equipment under extreme concurrent conditions.

[0028] S4. After the emergency is handled, control the robotic arm 21 to continue executing the intrinsic task based on the breakpoint pose. In step S4, after the emergency event is handled, the controller retrieves the lockout pose from step S2, plans a smooth backtracking trajectory, drives the robotic arm 21 to reset to the spatial state before the interruption, and reloads the remaining instruction stream of the intrinsic task. This achieves seamless breakpoint continuation of complex task flows, ensuring that the first priority routine production line tasks will not have missed actions or repeated grabbing due to repeated queueing of high priority exceptions.

[0029] S5. Evaluate the cycle deviation rate caused by handling the sudden event, and trigger motion parameter compensation of the robotic arm 21 or coordinated frequency reduction of the external feeding device 11 and / or external discharging device 12 based on the cycle deviation rate.

[0030] In step S5, the delay time caused by a series of correction actions such as preemption, track change, and reset is calculated by comparing the global clock, and the cycle deviation rate is obtained. The cycle deviation rate is used to perform self-compensation of motion parameters for the robotic arm 21, and instructions are directly sent to the PLC of the external device to reduce the stepping frequency of the conveyor belt.

[0031] Example 2 This embodiment provides a correction method for multi-process robotic arm 21 identification, which is applied to a multi-process production line system. (See attached document.) Figure 1As shown, the production line system in this embodiment includes at least an external feeding device 11, an external discharging device 12, a robotic arm 21, and a system controller for coordinating the operation of the entire workstation. To execute the correction method described in this embodiment, the controller needs to acquire multi-source data to construct a global state matrix. Combined with... Figure 1 As shown, the specific hardware acquisition link mapping for multi-source data is as follows: The production line visual image is obtained by capturing environmental images containing material shape and location information through a production line visual sensor 31 installed above the production line gantry rail or facing the work area.

[0032] The sensor detection signal originates from a physical detection component installed on the side / end of the external feeding device 11 or the external discharging device 12. When the material passes through a specific station, a level transition is triggered, generating the sensor detection signal representing the material's arrival status.

[0033] The encoder readings of the robotic arm 21 are derived from the servo drive encoders built into each rotary joint or linear motion module of the robotic arm 21, and are used to provide real-time feedback on the current physical deflection angle or displacement of each joint of the robotic arm 21.

[0034] The system clock data is provided by the underlying system of the control box 40 on the production line. The control box 40 runs an industrial-grade real-time operating system 34 or integrates a high-precision hardware clock source. The system clock data serves as a globally unified time reference, uniformly stamping the heterogeneous data asynchronously acquired by the vision cameras, sensors, and encoders with microsecond-level timestamps, thereby ensuring the time consistency of the data in the global state matrix.

[0035] In this embodiment, the multi-source data includes production line visual images, sensor detection signals, encoder readings of the robotic arm 21, and system clock data. The global state matrix is ​​a data set containing three-dimensional coordinates of the production line space, material attribute status, and timestamp information.

[0036] It should be noted that, since the sampling frequency and communication cycle of the vision camera, photoelectric sensor 32 and servo encoder 33 are different, it is easy to produce asynchronous problems where the material has already flowed away when the defective product is seen. The system clock data is introduced as a unified reference beat to give each frame of heterogeneous data a unified timestamp, thus ensuring the time reference.

[0037] In this embodiment, the step of acquiring multi-source data to construct a global state matrix includes: Based on the system clock data, timestamps are uniformly assigned to the real-time acquired production line visual images, sensor detection signals, and encoder readings of the robotic arm 21. Specifically, this ensures that the data extracted by the subsequent system at any time slice maintains a reliable time reference.

[0038] The visual images of the production line with the same timestamp and the sensor detection signals are analyzed to determine the material attribute status. Combined with the preset equipment installation calibration parameters, the material attribute status is mapped to the three-dimensional spatial coordinates of the material in the production line coordinate system. Specifically, visual images with unified timestamps are cross-validated with sensor signals to accurately determine whether the material is in a good, defective, or foreign object intrusion state. Then, the device installation calibration parameters that have been pre-calibrated and stored in the controller are called to map the material attribute state from the two-dimensional pixel coordinate system or the one-dimensional sensor position to the three-dimensional spatial coordinates (X, Y, Z) in the global base coordinate system of the production line.

[0039] The encoder readings of the robotic arm 21 body with the same timestamp are obtained, and spatial pose mapping is performed in combination with the physical link parameters of the robotic arm 21 to output the real-time three-dimensional spatial coordinates of the end effector of the robotic arm 21. Specifically, in the prior art, the position of the robotic arm 21 is purely tracked by external vision. In this embodiment, the original rotation data of the body servo encoder 33 with the same timestamp is directly read and combined with the inherent physical link parameters of the robotic arm 21 to perform forward kinematic spatial pose mapping, and deduce the current coordinates of the end flange or fixture of the robotic arm 21 in three-dimensional space. This avoids the occlusion blind spots and image processing delay problems of external vision tracking.

[0040] The global state matrix is ​​generated by combining the material attribute status, the spatial three-dimensional coordinates of the material, and the real-time spatial three-dimensional coordinates of the end effector of the robotic arm 21 at the same timestamp.

[0041] Specifically, using a unified timestamp as the index key, the calculated external environment data and the calculated self-entity data are assembled to generate a data structure containing dynamic elements of the production line panorama, namely a global state matrix. This allows for the retrieval of all decision-making basis with minimal computing power when performing anomaly preemption assessment, collision limit judgment, and cycle time compensation. This improves responsiveness.

[0042] In this embodiment, the step of controlling the robotic arm 21 to perform a first-priority intrinsic task along a preset path based on the global state matrix includes: Extract the three-dimensional spatial coordinates of the target material from the global state matrix as the target grasping pose; The standard working trajectory connecting the current position and the target grasping posture is generated by combining the real-time three-dimensional spatial coordinates of the end effector of the robotic arm 21 as the preset path; A drive command is sent to the servo controller of the robotic arm 21 to drive the robotic arm 21 to move along the preset path toward the target grasping pose.

[0043] It should be noted that the three-dimensional coordinates of the target material in the production line base coordinate system are extracted from the pre-constructed, real-time updated global state matrix and directly defined as the physical target point that the end effector of the robotic arm 21 needs to reach, i.e., the target grasping pose. The real-time three-dimensional coordinates of the end effector of the robotic arm 21 in the global state matrix are used as the dynamic starting point, and the above-mentioned target grasping pose is used as the ending point. The path planning algorithm is called to fit a non-interference coherent geometric trajectory in three-dimensional space, i.e., the standard operating trajectory. The planned standard operating trajectory is converted into a motor drive command sequence and sent to the servo controllers of each joint of the robotic arm 21, which drives the motors of each physical joint to operate at a predetermined speed and acceleration in a closed loop.

[0044] In this embodiment, the step of identifying sudden events based on the global state matrix includes: The material attribute states in the global state matrix are compared with a preset standard feature set in real time. If the material properties are found to meet the characteristics of high-risk defective products or foreign object intrusion, an abnormal signal is triggered. An abnormal state containing the characteristics of high-risk defective products or the characteristics of foreign object intrusion is identified as the sudden event.

[0045] It should be noted that the material attribute states abstracted from the matrix are cross-compared in real time with the standard feature set pre-stored in the controller's memory. When the comparison result hits these high-risk thresholds, an abnormal signal with extremely high interruption weight is triggered in the control bus. The stringent threshold conditions for abnormal triggering are clearly defined, effectively shielding low-priority routine defects in the production line, such as minor dimensional deviations and slight color differences, which are non-fatal defects that can be handled by subsequent workstations. This avoids invalid downtime caused by frequent interruptions of the robotic arm 21 due to minor abnormalities, ensuring the continuous and efficient operation of the main production line. Through a strict comparison and judgment mechanism, the objectively existing abnormal physical state is encapsulated into a sudden event with extremely high scheduling weight.

[0046] Example 3 This embodiment provides a further technical solution based on Embodiment 1 or Embodiment 2.

[0047] In this embodiment, the step of recording the current breakpoint pose includes: extracting the real-time physical deflection angle parameters of each joint of the robotic arm 21 at the current moment, and the real-time three-dimensional spatial coordinates and three-dimensional attitude angles of the end effector of the robotic arm 21 in the production line base coordinate system, and storing and protecting them as the breakpoint pose; The step of generating the target instruction for handling the emergency includes: extracting the real-time spatial three-dimensional coordinates and entity contour normal vector of the object corresponding to the emergency in the global state matrix; calculating the intervention angle without physical interference based on the entity contour normal vector as the desired end attitude target; and combining the intervention angle with the spatial target coordinates to generate the target instruction.

[0048] It should be noted that, at the millisecond instant of triggering the preemptive interrupt, the absolute / incremental encoder values ​​of each axis motor at the servo's underlying layer are directly read to obtain the most original real-time physical deflection angles of each joint. Simultaneously, the absolute three-dimensional coordinates (X, Y, Z) and three-dimensional attitude angles of the end effector of the robotic arm 21, calculated based on forward kinematics, are extracted and written into a protected non-volatile storage area or high-priority cache. Existing technology often results in the robotic arm 21 losing its current motion data during interruptions; for example, if it is holding a good product and moving it halfway, a reset can easily lead to a logic deadlock or collision. This step double-locks the internal joint states and external spatial attitude of the robotic arm 21, ensuring that any minute physical state at the moment of interruption is completely preserved, providing a geometric and electrical control reference for a lossless and smooth reset in subsequent steps. The system extracts not only the 3D center coordinates of the abnormal object from the global state matrix, but also the solid contour normal vectors of its surface or edges. Using these normal vectors as constraints, and combining them with the geometric interference model of the robotic arm 21's end effector, it calculates the optimal approximation angle that effectively grasps and touches the target while avoiding interference from surrounding material trays or conveyor belts—the desired end effector posture target. It also calculates the safest lower gripper angle. This improves the success rate of abnormal intervention and effectively protects the end effector and surrounding production line hardware from mechanical pressure.

[0049] In this embodiment, a further technical solution is provided, wherein the step of recording the current breakpoint pose further includes: Upon recognizing the sudden event, the shortest execution node required for the robotic arm 21 to safely place the currently held material is calculated in real time. The shortest execution node is used as a new interruptible node. After the robotic arm 21 moves to the new interruptible node, the real-time physical space state of the robotic arm 21 is extracted and protected as the breakpoint pose.

[0050] It should be noted that when the highest priority interrupt command is issued, the first step is to determine the current working state of the end effector of the robotic arm 21, such as whether it is in the middle of moving a good product. Combining the current motion velocity vector, acceleration limit, and the spatial structure of the surrounding material tray, the system deduces along the current preset path forward or towards the nearest temporary storage area to find the nearest spatial coordinate that meets gravity balance and ensures that the currently held material will not fall or be damaged due to a sudden stop. The shortest execution node is calculated. The shortest execution node calculated above is directly refreshed as the new interruptible node, and the action is given absolute local priority. The servo motors of each joint of the robotic arm 21 are controlled to smoothly finish the current action and move to the node, completing the safe placement of the material or the safe docking of the robotic arm itself. After confirming that the robotic arm 21 has entered a completely stationary steady state, the underlying encoder and coordinate reading command are triggered to seal the joint angles and spatial attitude of the real-time spatial state of the safe position as the breakpoint pose. By calculating the shortest execution node, it is equivalent to planning the nearest safe docking station for the high-speed robotic arm 21, ensuring the safety of the current semi-finished products and equipment hardware.

[0051] In this embodiment, the step of planning the motion trajectory according to the target instruction includes: A kinematic mapping model is constructed based on the solid link parameters of the robotic arm 21; The end-effector posture target and spatial target coordinates in the target command are input into the kinematic mapping model for inverse pose mapping, and the expected physical deflection angle sequence of each joint is output to fit and generate the motion trajectory.

[0052] It should be noted that by reading the physical link parameters corresponding to the robotic arm model 21, such as the physical length of each arm span, the physical offset distance between joints, and the link torsion angle, a forward and inverse kinematic mapping model corresponding to the real physical robotic arm 211:1 is constructed.

[0053] In this embodiment, the step of adjusting the end effector posture in the target command and replanning the motion trajectory if the expected motion parameters exceed the physical limits of the robotic arm 21, so as to drive the robotic arm 21 to handle the sudden event with the new motion trajectory, includes: When any joint angle in the expected physical deflection angle sequence exceeds the maximum mechanical deflection limit of that joint, the three-dimensional attitude angle corresponding to the expected end attitude target in the target command is compensated and adjusted. Specifically, when a dead angle exists that exceeds the factory settings or physical anti-collision boundaries of the servo axis, the system actively freezes the violating command and breaks the absolute rigid constraint of the original target command. While prioritizing accurate spatial coordinate reaching the target position, it proactively injects fine-tuning compensation into the end-effector's attitude target. By changing the tilt angle of the gripping or intervention, the system obtains the solution space for the current joint's mechanical dead angle, avoiding servo overload alarms and hard limit collisions that are easily triggered when the robotic arm 21 undergoes sudden large-span trajectory changes. This step utilizes the redundant degrees of freedom of the robotic arm 21, sacrificing minor non-critical attitude accuracy to ensure the core gripping position is reached, thus improving the equipment's robustness in extreme working conditions.

[0054] The adjusted end-effector pose target is re-input into the kinematic mapping model for inverse pose mapping until all joint angles are within the mechanical deflection limit, generating the compliant new motion trajectory. Specifically, the compensated new 3D attitude angles are recombine with the original spatial coordinates and input into the mapping model again for inverse calculation. If the calculated joint angles still exceed the limits, the compensation adjustment is accumulated and the solution is iterated again until all six or more joint angles output by the model strictly fall within the safe range of the maximum mechanical deflection limit, thereby fitting and generating a completely new obstacle avoidance trajectory. The system can quickly lock in a safe trajectory change path through self-iterative trial and error without any manual teaching or intervention.

[0055] The new motion trajectory joint control command sequence is sent to the servo controller of the robotic arm 21.

[0056] Specifically, the controller compiles the above-mentioned continuous joint angle sequence, which has been cross-validated by physical limits and confirmed to be absolutely safe, into position loop / velocity loop pulse control instructions that can be directly executed by the servo driver, and officially sends them to each underlying hardware node of the robotic arm 21 to drive the motor to operate.

[0057] In this embodiment, a further technical solution is provided, wherein the step of compensating and adjusting the three-dimensional attitude angle corresponding to the desired end-effector attitude target in the target instruction includes: Freeze erroneous joint control commands that exceed the maximum mechanical deflection limit; With the spatial position coordinates of the end effector remaining unchanged as a constraint, the roll angle α, pitch angle β, and yaw angle γ in the desired end attitude target are extracted. According to the preset attitude tolerance priority, the roll angle α, the pitch angle β and the yaw angle γ are sequentially injected with a preset step size of fine adjustment compensation amount, and the attitude adaptive reconstruction closed loop is executed until all calculated joint angles meet the mechanical deflection limit.

[0058] It should be noted that when an out-of-limit interrupt exceeding the mechanical deflection limit is triggered during the calculation process, the controller immediately intercepts the illegal instruction stream in the system memory buffer layer, preventing it from entering the position or velocity loop of the underlying servo driver. This ensures that fatal instructions that could cause interference, folding lock-up, or servo motor overload and burnout are prevented, avoiding production line shutdowns due to hardware alarms. By presetting priorities and fixed step sizes, computational power divergence or solution disasters caused by simultaneous adjustment of multiple variables are avoided, ensuring response speed; furthermore, fine-tuning according to priority ensures the safe posture obtained in the final solution.

[0059] Example 4 This embodiment provides a further technical solution based on Embodiment 1, Embodiment 2, or Embodiment 3.

[0060] In this embodiment, the step of controlling the robotic arm 21 to continue executing the intrinsic task based on the breakpoint pose includes: Extract the stored breakpoint pose; Specifically, the breakpoint data of the robotic arm 21, which was sealed during the preemption phase in step S2, is read from the high-priority cache or the protected non-volatile storage area.

[0061] Plan a smooth obstacle avoidance backtracking trajectory from the current pose at the end of the handling of the emergency to the breakpoint pose; Specifically, taking the current pose of the robotic arm 21 after the emergency is handled as the starting point and the extracted breakpoint pose as the ending point, and combining it with the latest environmental topology in the global state matrix, a continuous, abrupt, and interference-free backtracking trajectory is fitted in three-dimensional space using polynomial interpolation or spline curve algorithms. Existing technologies often employ rigid linear or joint interpolation for resetting, which is highly prone to collisions with equipment along the way. This step emphasizes smoothness and obstacle avoidance, protecting the spatial safety of the robotic arm 21 and peripheral production line equipment, and avoiding servo motor current surges caused by sudden acceleration changes, effectively extending the mechanical lifespan of the underlying hardware.

[0062] After driving the robotic arm 21 to reset to the breakpoint pose along the smooth obstacle avoidance retreat trajectory, the remaining execution instructions of the intrinsic task before the termination are retrieved and executed along the preset path.

[0063] Specifically, the controller drives each joint motor to move along the retraction trajectory. When the error between the real-time pose fed back by the encoder and the breakpoint pose approaches the zero tolerance band, the physical reset is confirmed to be successful. At this time, the system immediately unlocks the instruction stack of the intrinsic task and pushes the remaining execution instructions that were previously suspended due to preemption back into the underlying control loop, continuing to run along the original preset path generated in stage S1. This ensures that after the system has experienced extremely complex sudden track changes and anomaly handling, the original first-priority workflow has not experienced any breaks, omissions, or duplicate captures.

[0064] In this embodiment, the step of evaluating the cycle deviation rate caused by handling the sudden event, and triggering motion parameter compensation of the robotic arm 21 or coordinated frequency reduction of the external feeding device 11 and / or external discharging device 12 based on the cycle deviation rate includes: The additional time overhead caused by task interruption, emergency handling and reset process is statistically analyzed, and the cycle deviation rate of the current production line cycle time lagging behind the preset steady-state cycle time is calculated. Specifically, the accumulated additional time cost is compared with the standard production cycle originally set for the production line, and the cycle deviation rate, a parameter representing the degree of lag, is calculated.

[0065] When the beat deviation rate is less than or equal to a preset coordination threshold, the joint drive speed of the robotic arm 21 is increased in subsequent work cycles to perform the motion parameter compensation. Specifically, when the assessed cycle deviation rate is small, it means that the lag is still within the redundancy range of the physical acceleration capability of the servo motor of the robotic arm 21 and has not exceeded the coordination threshold. The controller does not require external equipment to directly modify the speed and acceleration feedforward parameters of the robotic arm 21 in the subsequent regular operation cycle, and gradually compensates for and eliminates the time lost in the previous cycle in the subsequent cycles. This avoids frequent intervention in the upstream and downstream conveyor belts due to minor abnormal interference, and utilizes the potential and operating efficiency of the robotic arm 21 itself.

[0066] When the cycle deviation rate is greater than the coordination threshold, a speed reduction control signal is sent to the external feeding device 11 and / or the external discharging device 12 to synchronously reduce the running step cycle of the external material conveying mechanism until the cycle deviation rate returns to the preset error range.

[0067] Specifically, when the deviation rate exceeds the physical limit of single-machine acceleration, that is, when the robotic arm 21 cannot compensate for the time no matter how much it accelerates, the control center immediately takes over the global bus and sends hard deceleration or delayed step instructions to the PLCs of external conveying equipment such as the blank feeding line and the finished product discharging line through the communication interface. This lengthens the input and output time interval of external materials and actively slows down the rhythm of the entire line to wait for the robotic arm 21, thus avoiding the chain reaction problem of global material accumulation and squeezing collision caused by local single-machine strikes / delays in traditional rigid production lines.

[0068] In this embodiment, the collaborative threshold includes a first threshold and a second threshold, the first threshold being 5% and the second threshold being 10%. The step of triggering motion parameter compensation of the robotic arm 21 based on the cycle deviation rate, or the collaborative frequency reduction of the external feeding device 11 and / or the external discharging device 12, includes: When the cycle deviation rate is less than or equal to 5%, the joint drive speed of the robotic arm 21 is only slightly adjusted in the subsequent work cycle, and the motion inflection points of the subsequent preset path are reduced. Specifically, when the lag amount is extremely small and the cycle deviation rate is less than or equal to 5%, it is determined that the deviation is completely within the normal torque redundancy of the 21 servo motors of a single robotic arm. The controller only issues a fine-tuning acceleration command in the next work cycle. At the same time, the trajectory smoothing algorithm is actively called to reduce unnecessary geometric motion inflection points in the subsequent preset path and replace right-angle pauses with spatial arcs.

[0069] When the cycle deviation rate is greater than 5% and less than or equal to 10%, the joint drive speed of the robotic arm 21 is continuously adjusted in the next two work cycles. Specifically, when the lag reaches a moderate level, forcibly accelerating to catch up within a single cycle will pose a serious risk of overload to the servo motor and cause severe vibration of the robotic arm 21. Therefore, the system introduces a time-domain smoothing strategy, which proportionally divides the total cycle time catch-up amount and distributes it over the next two consecutive work cycles using a stepped or gradual acceleration curve.

[0070] When the cycle deviation rate is greater than 10%, the coordinated frequency reduction of the external feeding device 11 and / or the external discharging device 12 is triggered. Specifically, when the deviation rate exceeds the limit of a single machine, it is determined that the robotic arm 21 can no longer compensate for the huge time difference by its own acceleration. It is necessary to immediately switch from internal drive to external intervention and accurately calculate the physical delay time based on the deviation rate. Subsequently, the delay parameter is issued through the PLC to directly overwrite the operating frequency of the underlying drive motor of the external conveyor belt, or modify its start-stop step time interval parameter.

[0071] The steps of the coordinated frequency reduction include: calculating the absolute delay time based on the cycle deviation rate, issuing a frequency reduction command containing the absolute delay time to the programmable logic controller of the external feeding device 11 and / or the external discharging device 12, proportionally reducing the operating frequency of the drive motor of the external material conveyor belt, or extending the start-stop step interval of the external material conveyor belt, so that the material input or output rate within the preset cycle matches the current actual processing cycle of the robotic arm 21.

[0072] Specifically, the frequency of the drive motor is proportionally reduced to adapt to the continuously operating variable frequency conveyor belt, and the start-stop step interval is extended to adapt to stepper or indexing conveyor lines. Under extreme and sudden working conditions, the input / output rate of external materials is matched to the actual processing cycle of the severely damaged robotic arm 21. This avoids the chain reaction problems caused by single-machine delays, such as severe accumulation of upstream materials, mold crushing and impact, and dry running of downstream equipment.

[0073] Example 5 See Figure 6 As shown, this embodiment provides a correction device for multi-process robotic arm recognition, used to apply the correction method for multi-process robotic arm recognition described in any of the above embodiments. The device includes: The matrix construction and execution module 100 is used to acquire multi-source data to construct a global state matrix, and control the robotic arm to execute a first-priority intrinsic task along a preset path based on the global state matrix; The breakpoint protection module 200 is used to identify sudden events based on the global state matrix during the execution of the intrinsic task. If the sudden event is evaluated as having a second priority and higher than the first priority, the current breakpoint pose is recorded and a target instruction for handling the sudden event is generated. The attitude reconstruction and trajectory change module 300 is used to adjust the end attitude in the target instruction and replan the motion trajectory when the expected motion parameters exceed the physical limits of the robotic arm, so as to drive the robotic arm to handle the sudden event with the new motion trajectory when the motion trajectory is planned according to the target instruction. The breakpoint continuation module 400 is used to control the robotic arm to continue executing the intrinsic task based on the breakpoint pose after the emergency event is handled. The cycle time compensation module 500 is used to evaluate the cycle time deviation rate caused by handling the sudden event, and trigger the motion parameter compensation of the robotic arm or the coordinated frequency reduction of the external feeding device and / or the external discharging device according to the cycle time deviation rate.

[0074] The multi-process robotic arm identification and correction device provided in this embodiment has the following advantages: By recording the current breakpoint pose and generating target instructions in real time when a high-priority emergency event is detected, preemptive intervention for high-risk anomalies is achieved. Simultaneously, based on the stored breakpoint pose, the robotic arm can be controlled to reset without damage and continue executing its intrinsic task after the emergency event is handled. This avoids the missed detection of high-risk defective products and solves the problems of workflow disruption, logic deadlock, and missed actions caused by traditional hard interruptions. When the robotic arm performs a large-span temporary trajectory change in response to an emergency event, if the expected motion parameters exceed physical limits, the end effector posture is actively adjusted to replan a compliant trajectory, avoiding servo limit alarms, overload shutdowns, or collisions caused by forced trajectory changes, thus improving the continuous operation rate and equipment safety of the robotic arm in complex concurrent environments. In addition, by assessing and handling the cycle deviation rate caused by emergencies, it can not only trigger the robot arm's own motion parameter compensation, but also trigger the frequency reduction of external feeding / discharging equipment when the delay is severe, thus solving the problem of chain reactions such as severe accumulation and compression of upstream and downstream materials caused by abnormal time consumption of a single device.

[0075] Figure 7 An internal structural diagram of a computer device in one embodiment is shown. This computer device can specifically be a terminal or a server. Figure 7 As shown, the computer device includes a processor, memory, and network interface connected via a system bus. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and may also store a computer program. When executed by the processor, this computer program enables the processor to implement a correction method for multi-process robotic arm recognition. The memory may also store a computer program, which, when executed by the processor, enables the processor to implement a correction method for multi-process robotic arm recognition. Those skilled in the art will understand that... Figure 7 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0076] It should be noted that the structures, proportions, sizes, etc., shown in the accompanying drawings of this specification are only for the purpose of assisting those skilled in the art in understanding and reading the content disclosed in the specification, and are not intended to limit the conditions under which this application can be implemented. Therefore, they have no substantial technical significance. Any modifications to the structure, changes in the proportions, or adjustments to the size should still fall within the scope of the technical content disclosed in this application, provided that they do not affect the effects and purposes that this application can produce.

[0077] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments described above. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM), etc.

[0078] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0079] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A method for correcting deviations in multi-process robotic arm identification, applied to a production line system, wherein the production line system includes at least a robotic arm, an external feeding device, an external discharging device, and a controller, characterized in that, The method is executed by the controller and includes the following steps: Acquire multi-source data to construct a global state matrix, and control the robotic arm to execute a first-priority intrinsic task along a preset path based on the global state matrix; During the execution of the intrinsic task, a sudden event is identified based on the global state matrix. If the sudden event is evaluated as having a second priority and higher than the first priority, the current breakpoint pose is recorded and a target instruction for handling the sudden event is generated. When planning the motion trajectory according to the target instruction, if the expected motion parameters exceed the physical limits of the robotic arm, the end-effector posture in the target instruction is adjusted and the motion trajectory is replanned to drive the robotic arm to handle the sudden event with the new motion trajectory. After the emergency is handled, the robotic arm is controlled to continue executing the intrinsic task based on the breakpoint pose. Assess the cycle deviation rate caused by handling the sudden event, and trigger motion parameter compensation of the robotic arm or coordinated frequency reduction of the external feeding device and / or external discharging device based on the cycle deviation rate.

2. The correction method for multi-process robotic arm recognition according to claim 1, characterized in that, The multi-source data includes production line visual images, sensor detection signals, robotic arm encoder readings, and system clock data. The global state matrix is ​​a data set containing three-dimensional coordinates of the production line space, material attribute status, and timestamp information.

3. The correction method for multi-process robotic arm recognition according to claim 2, characterized in that, The steps for acquiring multi-source data to construct a global state matrix include: Based on the system clock data, timestamps are uniformly assigned to the real-time acquired production line visual images, sensor detection signals, and encoder readings of the robotic arm body; The visual images of the production line with the same timestamp and the sensor detection signals are analyzed to determine the material attribute status. Combined with the preset equipment installation calibration parameters, the material attribute status is mapped to the three-dimensional spatial coordinates of the material in the production line coordinate system. The encoder readings of the robotic arm body with the same timestamp are obtained, and spatial pose mapping is performed in combination with the physical link parameters of the robotic arm to output the real-time three-dimensional spatial coordinates of the robotic arm end effector. The global state matrix is ​​generated by combining the material attribute status, the spatial three-dimensional coordinates of the material, and the real-time spatial three-dimensional coordinates of the robotic arm end effector at the same timestamp.

4. The correction method for multi-process robotic arm recognition according to claim 1, characterized in that, The steps of controlling the robotic arm to perform a first-priority intrinsic task along a preset path based on the global state matrix include: Extract the three-dimensional spatial coordinates of the target material from the global state matrix as the target grasping pose; The preset path is generated by combining the real-time three-dimensional spatial coordinates of the robotic arm's end effector to connect the current position with the target grasping posture. A drive command is sent to the robotic arm servo controller to drive the robotic arm to move along the preset path toward the target grasping pose.

5. The correction method for multi-process robotic arm recognition according to claim 1, characterized in that, The step of identifying sudden events based on the global state matrix includes: The material attribute states in the global state matrix are compared with a preset standard feature set in real time. If the material properties are found to meet the characteristics of high-risk defective products or foreign object intrusion, an abnormal signal is triggered. An abnormal state containing the characteristics of high-risk defective products or the characteristics of foreign object intrusion is identified as the sudden event.

6. The correction method for multi-process robotic arm recognition according to claim 1, characterized in that: The step of recording the current breakpoint pose includes: extracting the real-time physical deflection angle parameters of each joint of the robotic arm at the current moment, and the real-time three-dimensional spatial coordinates and three-dimensional attitude angles of the robotic arm end effector in the production line base coordinate system, and storing and protecting them as the breakpoint pose; The step of generating the target instruction for handling the emergency includes: extracting the real-time spatial three-dimensional coordinates and entity contour normal vector of the object corresponding to the emergency in the global state matrix; calculating the intervention angle without physical interference based on the entity contour normal vector as the desired end attitude target; and combining the intervention angle with the spatial target coordinates to generate the target instruction.

7. The correction method for multi-process robotic arm recognition according to claim 6, characterized in that, The step of planning the motion trajectory according to the target instruction includes: A kinematic mapping model is constructed based on the physical link parameters of the robotic arm; The end-effector posture target and spatial target coordinates in the target command are input into the kinematic mapping model for inverse pose mapping, and the expected physical deflection angle sequence of each joint is output to fit and generate the motion trajectory.

8. The correction method for multi-process robotic arm recognition according to claim 7, characterized in that, The step of adjusting the end effector posture in the target command and replanning the motion trajectory if the expected motion parameters exceed the physical limits of the robotic arm, and driving the robotic arm to handle the sudden event with the new motion trajectory, includes: When any joint angle in the expected physical deflection angle sequence exceeds the maximum mechanical deflection limit of that joint, the three-dimensional attitude angle corresponding to the expected end attitude target in the target command is compensated and adjusted. The adjusted end-effector pose target is re-input into the kinematic mapping model for inverse pose mapping until all joint angles are within the mechanical deflection limit, generating the compliant new motion trajectory. The new motion trajectory joint control command sequence is sent to the robotic arm servo controller.

9. The correction method for multi-process robotic arm recognition according to claim 1, characterized in that, The step of controlling the robotic arm to continue executing the intrinsic task based on the breakpoint pose includes: Extract the stored breakpoint pose; Plan a smooth obstacle avoidance backtracking trajectory from the current pose at the end of the handling of the emergency to the breakpoint pose; After driving the robotic arm to reset to the breakpoint pose along the smooth obstacle avoidance retreat trajectory, the remaining execution instructions of the intrinsic task before the termination are retrieved and executed along the preset path.

10. The correction method for multi-process robotic arm recognition according to claim 1, characterized in that, The steps of assessing the cycle deviation rate caused by handling the sudden event, and triggering motion parameter compensation of the robotic arm based on the cycle deviation rate, or the coordinated frequency reduction of the external feeding device and / or external discharging device, include: The additional time overhead caused by task interruption, emergency handling and reset process is statistically analyzed, and the cycle deviation rate of the current production line cycle time lagging behind the preset steady-state cycle time is calculated. When the cycle deviation rate is less than or equal to a preset coordination threshold, the joint drive speed of the robotic arm is increased in subsequent work cycles to perform the motion parameter compensation. When the cycle deviation rate is greater than the coordination threshold, a speed reduction control signal is sent to the external feeding device and / or external discharging device to synchronously reduce the running step cycle of the external material conveying mechanism until the cycle deviation rate returns to the preset error range.

11. The correction method for multi-process robotic arm recognition according to claim 10, characterized in that, The collaborative threshold includes a first threshold and a second threshold, the first threshold being 5% and the second threshold being 10%. The step of triggering motion parameter compensation of the robotic arm based on the cycle deviation rate, or the collaborative frequency reduction of the external feeding device and / or the external discharging device, includes: When the cycle deviation rate is less than or equal to 5%, the joint drive speed of the robotic arm is only finely adjusted in the subsequent work cycle, and the motion inflection points of the subsequent preset path are reduced. When the cycle deviation rate is greater than 5% and less than or equal to 10%, the joint drive speed of the robotic arm is continuously adjusted in the next two work cycles. When the cycle deviation rate is greater than 10%, the coordinated frequency reduction of the external feeding equipment and / or the external discharging equipment is triggered. The steps of the coordinated frequency reduction include: calculating the absolute delay time based on the cycle deviation rate, issuing a frequency reduction command containing the absolute delay time to the programmable logic controller of the external feeding device and / or external discharging device, proportionally reducing the operating frequency of the drive motor of the external material conveyor belt, or extending the start-stop step interval of the external material conveyor belt, so that the material input or output rate within the preset cycle matches the current actual processing cycle of the robotic arm.

12. A correction device for multi-process robotic arm recognition, used in applying the correction method for multi-process robotic arm recognition according to any one of claims 1 to 11, characterized in that, The device includes: The matrix construction and execution module is used to acquire multi-source data to construct a global state matrix, and control the robotic arm to execute a first-priority intrinsic task along a preset path based on the global state matrix; The breakpoint protection module is used to identify sudden events based on the global state matrix during the execution of the intrinsic task. If the sudden event is evaluated as having a second priority and higher than the first priority, the current breakpoint pose is recorded and a target instruction for handling the sudden event is generated. The attitude reconstruction and trajectory change module is used to adjust the end attitude in the target instruction and replan the motion trajectory when the expected motion parameters exceed the physical limits of the robotic arm, so as to drive the robotic arm to handle the sudden event with the new motion trajectory when the motion trajectory is planned according to the target instruction. The breakpoint resume module is used to control the robotic arm to continue executing the intrinsic task based on the breakpoint pose after the emergency event has been handled. The cycle time compensation module is used to evaluate the cycle time deviation rate caused by handling the sudden event, and trigger the motion parameter compensation of the robotic arm or the coordinated frequency reduction of the external feeding device and / or the external discharging device based on the cycle time deviation rate.

Images