Teleoperation control method, equipment and system thereof

Abstract

According to the teleoperation control method, device and system, in a first control mode, on the basis of joint movement signals from an isomorphic main mechanical arm, the operation requirements of rapid positioning or moving and the like of a slave mechanical arm in a large range are met through isomorphic mapping. And in the second control mode, based on an increment pose signal for the tail end of the slave mechanical arm, the requirement for fine and smooth operation of the tail end of the slave mechanical arm in the close contact stage with the operation object is met, and seamless and safe switching between the joint space isomorphic mapping mode and the tail end increment control mode of the mechanical arm is controlled, so that the joint space isomorphic mapping and the tail end increment control are achieved. And the requirements of speed and precision in teleoperation can be met. And meanwhile, the control mode is switched according to the mode switching trigger instruction generated in the current operation stage of the slave mechanical arm, the operation requirements of the slave mechanical arm in different operation scenes are met, the smoothness of the operation process is guaranteed, and high-quality demonstration data are provided for the subsequent autonomous learning and intelligent decision-making capacity of the robot.

Description

Technical Field

[0001] This application relates to the field of robot control technology, and in particular to a remote operation control method, device and system thereof. Background Technology

[0002] With the rapid development of intelligent manufacturing and flexible automation technologies, the demand for automation in fabric sewing processes is increasing, especially in labor-intensive industries such as clothing and home textiles. Achieving high-precision and highly adaptable fabric handling has become a key link in promoting industrial upgrading. In the process of building robotic intelligent systems for sewing scenarios, high-quality operational data acquisition is the fundamental support for training and optimizing artificial intelligence models. Teleoperation technology, as an effective means of human-machine collaboration, can provide robots with realistic and reliable motion and force data by having human operators demonstrate complex and dexterous movements.

[0003] However, current teleoperation systems face key technical bottlenecks in applications requiring precise control and multimodal perception, such as fabric handling. For example, current systems often use a master-end actuator to directly map the slave gripper's movements to grip and transfer the fabric. However, this approach struggles to balance large-scale, rapid movement with precise adjustments during the near-contact phase. This is because high-speed movement is prone to vibration or positioning errors, while the system lacks sufficient compliance and adaptability at the moment of contact. Consequently, the slave arm experiences stuttering and shaking during execution, affecting operational smoothness, severely limiting the acquisition of high-quality demonstration data, and ultimately impacting the robot's subsequent autonomous learning and intelligent decision-making capabilities. Summary of the Invention

[0004] To overcome the problems existing in related technologies, this manual provides remote operation control methods, equipment and systems.

[0005] Firstly, a teleoperation control method is provided, the method comprising: Receive a mode switching trigger instruction, which is generated based on the current operation stage of the robotic arm, and the current operation stage is used to characterize different operational requirements for the robotic arm; In response to a mode switching trigger command, the control mode of the slave robotic arm is switched to the target control mode, and the movement of the slave robotic arm is controlled based on the target control mode; The target control mode is either a first control mode or a second control mode. In the first control mode, the slave robot arm is controlled to perform joint space mapping motion based on the joint motion signal from the isomorphic master robot arm. In the second control mode, the slave robot arm is controlled to perform end-effector incremental motion based on the incremental pose signal for the end effector of the slave robot arm.

[0006] According to a teleoperation control method provided in this application, the current operation stage includes a first operation stage in which the slave robot arm and the work object are in a non-contact state, and the step of switching the control mode of the slave robot arm to the target control mode in response to a mode switching trigger command includes: In response to the mode switching trigger command, the control mode of the robotic arm is switched to the first control mode.

[0007] According to a teleoperation control method provided in this application, the current operation stage includes a second operation stage in which the slave robot arm and the work object are in contact or nearly in contact. The step of switching the control mode of the slave robot arm to the target control mode in response to a mode switching trigger command includes: In response to the mode switching trigger command, the control mode of the robotic arm is switched to the second control mode.

[0008] According to the teleoperation control method provided in this application, the generation method of the mode switching trigger command includes: Obtain the relative position information and / or contact force information between the robotic arm and the work object; When the relative position information and / or contact force information satisfy the first condition characterizing that the current operation stage of the robotic arm is in the second operation stage, a mode switching trigger command to switch to the second control mode is generated. The first condition includes: the relative position information is less than the safe position threshold, and / or the contact force information is greater than zero.

[0009] According to the teleoperation control method provided in this application, the generation method of the mode switching trigger command includes: In response to a mode switching trigger request input by the user based on the current operation stage of the robotic arm, a corresponding mode switching trigger instruction is generated based on the mode switching trigger request.

[0010] According to a teleoperation control method provided in this application, switching the control mode of the robotic arm to a target control mode includes: Based on the actual end-effector pose at the moment of switching from the robotic arm, a motion trajectory is generated that smoothly transitions to the initial desired end-effector pose in the target control mode. Control commands are generated based on the motion trajectory to control the robotic arm to move along the motion trajectory.

[0011] According to a teleoperation control method provided in this application, before switching the control mode of the robotic arm to the target control mode, the method further includes: Stop instruction transmission corresponding to the current control mode and clear any unexecuted instructions from the instruction buffer.

[0012] According to a teleoperation control method provided in this application, when the target control mode is the second control mode, the robotic arm is moved based on the target control mode. The method further includes: Activate the signal input channel corresponding to the second control mode; The incremental pose signal received through the signal input channel for the robotic arm end effector is superimposed with the current actual pose of the end effector to obtain the desired end effector pose. Based on the desired end pose, the target value of the joint space is obtained by inverse kinematics solution; The target value of the joint space is sent to the servo system of the slave robot to control the slave robot to perform end effector incremental motion.

[0013] According to a teleoperation control method provided in this application, after activating the signal input channel corresponding to the second control mode, the method further includes: Reliability processing is performed on the incremental pose signal received through the signal input channel for the robotic arm end effector. The processed incremental pose signal is superimposed on the actual pose of the current end effector to obtain the desired end effector pose.

[0014] According to a teleoperation control method provided in this application, the method further includes: Obtain the contact force information between the end effector of the robotic arm and the work object; Based on the contact force information, the end-effector motion compensation amount is obtained through the admittance control model; The end-effector motion compensation amount is used to correct the desired end-effector pose generated based on the pose increment signal; The movement of the robotic arm is controlled based on the corrected desired pose.

[0015] According to the teleoperation control method provided in this application, after acquiring the contact force information between the end effector of the robotic arm and the work object, the method further includes: When the contact force corresponding to the contact force information exceeds a preset safety threshold, the generation of the desired end pose based on the pose increment signal is paused. The robotic arm is controlled to execute a retreat trajectory generated based on the admittance control model.

[0016] According to a teleoperation control method provided in this application, when the target control mode is the first control mode, the step of controlling the movement of the robotic arm based on the target control mode includes: Obtain the joint angle data of the master robotic arm, which is isomorphic to the slave robotic arm; A smooth trajectory sequence is generated based on the joint angle data, the smooth trajectory sequence including multiple target joint angles arranged in chronological order; The smooth trajectory sequence is sent to the servo system of the slave robot to control the slave robot to perform motion tracking according to the target joint angle.

[0017] According to the teleoperation control method provided in this application, before acquiring the joint angle data of the master robotic arm, which is isomorphic to the slave robotic arm, the method further includes: A gravity-compensating current is applied to the main robotic arm to bring it into a state of force balance.

[0018] According to a teleoperation control method provided in this application, applying a gravity compensation current to the main robotic arm includes: Based on the isomorphic dynamic model of the main manipulator, the gravitational torque of the main manipulator in the current posture is determined; The gravitational torque is converted into gravity compensation current for the motors of each joint of the main robotic arm; The gravity compensation current is applied to each joint motor of the main robotic arm.

[0019] According to a teleoperation control method provided in this application, the method further includes: The dynamic model is optimized, and the gravitational torque of the main robotic arm in the current posture is determined based on the optimized dynamic model; The dynamic model is optimized through parameter identification, which is based on the motion data and current data of the main robotic arm under various excitation trajectories.

[0020] According to a teleoperation control method provided in this application, the method further includes: Obtain the real-time joint state information of the robotic arm; The real-time joint state information is mapped to the isomorphic main robotic arm, and the main robotic arm is driven to perform corresponding posture synchronization.

[0021] According to a teleoperation control method provided in this application, the method further includes: During the control of the slave robotic arm, a dataset for training the slave robotic arm to autonomously execute tasks is recorded simultaneously; The dataset includes at least the joint motion sequence, end-effector pose sequence, and contact force information sequence of the robotic arm, and the sequences are aligned by timestamps.

[0022] According to a second aspect of the embodiments of this specification, an apparatus is provided, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements any of the teleoperation control methods described above.

[0023] Thirdly, a teleoperation control system is provided, the system comprising a homogeneous master robotic arm and a slave robotic arm, a slave controller for tracking the motion of the slave robotic arm, and an electronic device for implementing any of the teleoperation control methods described above.

[0024] Fourthly, a computer-readable storage medium is provided, on which a teleoperation control program is stored, wherein the teleoperation control program, when executed, implements the steps of any of the teleoperation control methods described above.

[0025] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the teleoperation control method as described in any of the above.

[0026] This application provides a remote operation control method, device, and system, which have the following beneficial effects: Two control modes are implemented: a first control mode and a second control mode. In the first control mode, based on joint motion signals from the isomorphic master robotic arm, the slave robotic arm is controlled to perform joint space mapping motion. This isomorphic mapping enables the slave robotic arm to perform tasks such as rapid positioning or movement over a wide range. In the second control mode, based on incremental pose signals from the end effector of the slave robotic arm, the slave robotic arm is controlled to perform incremental end effector motion. This enables precise and compliant operation during the near-contact phase between the end effector and the work object. By seamlessly and safely switching between the joint space isomorphic mapping and incremental end effector control modes, the system can balance the speed and accuracy requirements of teleoperation. Simultaneously, a corresponding mode switching trigger command is generated when the current work stage of the slave robotic arm is determined. This command switches the control mode of the slave robotic arm to the mode corresponding to the current work stage, thus meeting the operational requirements of the slave robotic arm at that stage. The command switching ensures the smoothness of the operation process. In summary, the adaptive dual-mode control based on work stage perception meets the needs of different work scenarios and provides high-quality demonstration data for the robot's subsequent autonomous learning and intelligent decision-making capabilities.

[0027] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this specification. Attached Figure Description

[0028] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this specification and, together with the description, serve to explain the principles of this specification.

[0029] Figure 1 This is a schematic flowchart illustrating a teleoperation control method according to an exemplary embodiment of this specification; Figure 2 This is another schematic flowchart illustrating a teleoperation control method according to an exemplary embodiment of this specification; Figure 3 This is a schematic block diagram of a remote control device illustrated in this specification according to an exemplary embodiment. Detailed Implementation

[0030] The technical solutions in the embodiments (or "implementations") of this application will be clearly and completely described herein with reference to the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements.

[0031] If the embodiments of this application contain terms relating to directional indications or positional relationships (such as up, down, left, right, front, back, inside, outside, top, bottom, center, vertical, horizontal, longitudinal, transverse, length, width, counterclockwise, clockwise, axial, radial, circumferential, etc.), such terms are only used to explain the relative positional relationships and movement of the components in a specific posture (as shown in the attached figures); if the specific posture changes, the directional indications or positional relationships will also change accordingly. Furthermore, the terms "first" and "second" used in the embodiments of this application are only for descriptive convenience and should not be construed as indicating or implying relative importance.

[0032] This application provides a remote operation control method, device, and system. The following detailed description, in conjunction with the accompanying drawings, illustrates this application. The features described in the embodiments and implementations can be combined with each other.

[0033] To address the aforementioned technical problems, this specification provides a remote operation control method.

[0034] The system aims to construct an adaptive teleoperation system for application scenarios requiring precise control and multimodal perception. During the large-scale non-contact phase, it employs a full-dimensional mapping mode from the master robot's joint space to the slave robot's joint space, achieving efficient and intuitive motion reproduction. When entering the precision operation area, it switches to manual operation mode via a manual or automatic trigger mechanism. The operator inputs incremental pose commands from the robot's end-effector coordinate system, and the system generates corresponding joint movements using real-time inverse kinematics, thereby adjusting the end effector. This significantly improves the control accuracy and adaptability in complex operating environments involving objects with complex deformations.

[0035] It should be noted that the teleoperation control method of this application is aimed at application scenarios that require fine manipulation and multimodal perception, such as teleoperation of flexible objects, including but not limited to fabric separation, film processing, flexible circuit board assembly, etc.

[0036] In sewing scenarios, achieving high-precision and highly adaptable fabric manipulation has become a key aspect of driving industrial upgrading. However, due to various interference factors such as fabric adhesion, wrinkles, and static electricity, as well as the complex and varied deformation patterns, the typical pre-sewing task of separating single-layer fabric is difficult to achieve. Therefore, this paper will describe the teleoperation control method, equipment, and system using a teleoperated system in a sewing automation scenario as an example. The method is equally applicable to other similar scenarios, and the operation is similar, so it will not be elaborated further here.

[0037] This application provides an embodiment of a remote operation control system.

[0038] The teleoperation control system (hereinafter referred to as the system for ease of description) includes a master robotic arm and a slave robotic arm that are homogeneous, a slave controller that realizes motion tracking of the slave robotic arm, and electronic equipment that realizes the subordinate teleoperation control method.

[0039] The master robotic arm, serving as the master input device, is structurally isomorphic to the slave robotic arm. Its function is to be directly dragged by the operator, generating intuitive joint motion signals. These signals are used to map and control the movement of the slave robotic arm in the first control mode, while gravity compensation enhances the user experience.

[0040] In this embodiment, the main robotic arm is configured with a seven-degree-of-freedom isomorphic robotic arm. Other isomorphic robotic arms, such as six-degree-of-freedom robotic arms, are also applicable.

[0041] The robotic arm serves as the execution end, performing tasks remotely at the work site. Its end effector is equipped with a gripper and a six-dimensional force sensor to interact with the environment and provide feedback on contact information.

[0042] In this embodiment, a robotic arm performs a single-layer fabric separation task on a sewing workbench. This task involves a large-scale movement away from the target, requiring rapid response capabilities to improve work efficiency. Simultaneously, it involves approaching the fabric and performing critical actions such as gripping and layering, demanding extremely high spatial accuracy and force control compliance to prevent multi-layer adhesion, stretching deformation, or tearing. Subsequent remote control methods can effectively meet the operational requirements for performing single-layer fabric separation tasks.

[0043] The slave controller, acting as the underlying execution unit, is directly connected to the servo motors of each joint of the slave robotic arm. Its function is to receive joint angle commands from the master controller and perform high-frequency, high-precision position / current closed-loop control to ensure that the slave robotic arm accurately tracks the command trajectory.

[0044] The electronic device, also known as the master controller, serves as the core of the system's computation, running the teleoperation control program and implementing all high-level decisions and algorithms. This includes, but is not limited to, subsequent tasks such as operation phase determination, dual-mode switching decisions, trajectory planning, inverse kinematics solving, admittance control, gravity compensation calculation, data acquisition and synchronization, and coordinating command and status communication between the master and slave devices.

[0045] In this embodiment, the electronic device may be, but is not limited to, a master controller, a computer, an industrial control computer, etc.

[0046] The specific operation process is as follows: After the system is started, the main controller initializes each module, and the main robotic arm is in a hovering state that can be easily dragged under gravity compensation.

[0047] The operator can manually trigger the mode switch or the system can automatically trigger it based on task requirements (such as large-scale positioning or fine stripping).

[0048] If currently in the first control mode, the operator directly drags the main robotic arm. High-precision encoders integrated into each joint of the main robotic arm collect joint angles in real time. The main controller smooths these joint angles, generates a joint space trajectory sequence, and sends it to the slave controller. Enabling a full-dimensional mapping mode from the main robotic arm joint space to the slave robotic arm joint space, the slave controller drives each joint of the slave robotic arm to reproduce the movement of the main robotic arm, achieving fast and intuitive large-range positioning. Simultaneously, the actual joint state of the slave robotic arm is fed back in real time and drives the main robotic arm to synchronize, maintaining transparency.

[0049] If currently in the second control mode, the operator uses the manual control device to input the end effector pose increment command, and combines this with real-time inverse kinematics to generate the corresponding seven-degree-of-freedom joint motion, which is then sent to the slave controller for execution. This allows for the adjustment of the end effector, achieving high-precision and smooth operation. During this process, if the contact force between the end effector and the workpiece exceeds the limit, the system immediately suspends the manual control device commands and prioritizes admittance yielding motion to protect the fabric.

[0050] Through the aforementioned collaborative operation, efficient, precise, and safe remote operation was achieved for challenging tasks such as separating single-layer fabrics.

[0051] In some embodiments, throughout the control process, the master controller synchronously records multimodal data, including the joint angles, end-effector pose, and contact force information of the robotic arm. These data are aligned using timestamps to form a high-quality dataset for robot skill learning. In other words, in building a robotic intelligent system for sewing scenarios, this high-quality operational data provides fundamental support for training and optimizing artificial intelligence models.

[0052] The aforementioned remote operation control system meets the comprehensive requirements of high precision, high stability, and strong adaptability for fabric separation tasks in sewing scenarios. It not only helps improve the quality and efficiency of fabric operation data acquisition, but also lays a solid data and algorithm foundation for robots to autonomously complete complex sewing tasks in the future. It has important theoretical value and broad application prospects.

[0053] This application provides an embodiment of a teleoperation control method based on a teleoperation control system, wherein the teleoperation control method is applied to the master controller in the teleoperation control system.

[0054] Reference Figure 1 , Figure 1 This is a schematic flowchart of a remote operation control method provided in the embodiments of this specification.

[0055] Specifically, this includes the following steps 101 to 102: Step 101: Receive a mode switching trigger instruction, which is generated based on the current operation stage of the robotic arm, and the current operation stage is used to characterize different operational requirements for the robotic arm. Step 102: In response to the mode switching trigger command, the control mode of the slave robotic arm is switched to the target control mode, and the movement of the slave robotic arm is controlled based on the target control mode; The target control mode is either a first control mode or a second control mode. In the first control mode, the slave robot arm is controlled to perform joint space mapping motion based on the joint motion signal from the isomorphic master robot arm. In the second control mode, the slave robot arm is controlled to perform end-effector incremental motion based on the incremental pose signal for the end effector of the slave robot arm.

[0056] In this embodiment, the switching between the first control mode and the second control mode can be performed based on the current operation stage of the robotic arm. This includes switching from the current first control mode to the second control mode, and also switching from the current second control mode to the first control mode. In actual sewing scenarios, when the remote operation control system is activated, the robotic arm will first move rapidly over a large area to the vicinity of the fabric, and then perform actions such as grasping and layering. After performing these actions, it will again move rapidly over a large area to the area where the fabric is placed or from the initial position of the robotic arm.

[0057] Therefore, the following task execution sequence based on the sewing scenario will be explained using the current control mode as the first control mode as an example. The specific process is as follows: Currently, the current operation phase includes the first operation phase in which the robotic arm and the work object are in a non-contact state, indicating that the robotic arm needs to move rapidly over a wide range.

[0058] Reference Figure 2 , Figure 2 This is another schematic flowchart of a remote operation control method provided in the embodiments of this specification.

[0059] Step one involves controlling the movement of the slave robotic arm in the first control mode. This includes the acquisition and preprocessing of master robotic arm joint data, master-slave joint spatial mapping, and large-range rapid movement control.

[0060] In some embodiments, when the target control mode is the first control mode, controlling the movement of the robotic arm based on the target control mode includes the following steps a1 to a3: Step a1: Obtain the joint angle data of the master robotic arm, which is isomorphic to the slave robotic arm.

[0061] During the data acquisition and preprocessing process of the main robotic arm joints: The master end is equipped with a seven-DOF isomorphic robotic arm, including a master robotic arm and slave robotic arms. Each joint of the robotic arm integrates a high-precision encoder for real-time measurement of joint rotation angles.

[0062] First, the encoder acquires joint angles in real time at a set frequency.

[0063] To accurately and without distortion record the movements of the main robotic arm, a frequency of no less than 100Hz is set to ensure that rapid fine-tuning movements can be captured.

[0064] Next, the collected joint angle data is preprocessed, including the changes in the angles of each joint. .

[0065] In this embodiment, preprocessing includes, but is not limited to, median filtering and normalization.

[0066] As an example, a fixed-length data window is set for each joint angle data point. New data is added to the window each time it is acquired, while the oldest data is removed, keeping the window size constant. Then, all data within the window is sorted, and the median value is used as the filtered output for the current moment. Median filtering effectively eliminates transient, large-amplitude noise caused by electrical interference, encoder reading jumps, etc., preventing this noise from affecting the smoothness of subsequent motion.

[0067] As an example, the joint angle data after median filtering is normalized and mapped to a unified interval to eliminate the weight imbalance caused by differences in joint range of motion, serving as the basic input for subsequent interpolation and mapping.

[0068] Subsequently, the processed joint angle data is written to a thread-safe buffer. Subsequent master-slave mapping processes read the latest data from the buffer, avoiding direct access to the data being processed.

[0069] Step a2: Generate a smooth trajectory sequence based on the joint angle data. The smooth trajectory sequence includes multiple target joint angles arranged in chronological order.

[0070] Because operator input may be jittery, and direct mapping may lead to discontinuous arm movement, it is necessary to smooth the target joint angle.

[0071] In this embodiment, a sigmoid-like function with acceleration and velocity constraints and a linear interpolation algorithm are used to generate a smooth trajectory sequence based on the normalized joint angle data of the main robotic arm. The trajectory generation is achieved by limiting the maximum speed and acceleration to ensure smooth movement of the robotic arm and avoid impacts. The desired joint angle for each control cycle is calculated through interpolation to form a target joint angle sequence. The target joint angle refers to the expected angle value that each of the seven joints of the robotic arm needs to achieve at a specific moment.

[0072] Step a3: The smooth trajectory sequence is sent to the servo system of the slave robot to control the slave robot to perform motion tracking according to the target joint angle.

[0073] The smooth trajectory sequence is sent to the slave controller via a fixed baud rate communication link.

[0074] In this embodiment, the communication link includes a 1Mbps UART and a real-time Ethernet.

[0075] Receive target joint angle from robotic arm Then, the servo system of the robotic arm completes motion tracking, achieving high-fidelity reproduction of the main arm's movements. For example, based on the actual pose of the current end effector of the robotic arm and the target joint angle... The change in joint angle after forward mapping is obtained. The servo system of the robotic arm controls the movement of the robotic arm according to the changes in joint angles.

[0076] This mode is suitable for long-distance, rapid movement and has the advantages of fast response and low latency, significantly improving work efficiency.

[0077] In some embodiments, before acquiring joint angle data of the master robotic arm isomorphic to the slave robotic arm, the method further includes: A gravity-compensating current is applied to the main robotic arm to bring it into a state of force balance.

[0078] In the first control mode, the robotic arm requires the operator to continuously apply force to maintain its posture. However, prolonged operation can easily lead to fatigue, reducing data acquisition efficiency and quality. Therefore, to reduce the burden of human-machine interaction, the main robotic arm integrates a gravity compensation function, allowing the main robotic arm to hover in any posture without the operator needing to exert continuous force, thus significantly reducing fatigue.

[0079] In some embodiments, applying a gravity-compensating current to the main robotic arm includes: Based on the isomorphic dynamic model of the main manipulator, the gravitational torque of the main manipulator in the current posture is determined; The gravitational torque is converted into gravity compensation current for the motors of each joint of the main robotic arm; The gravity compensation current is applied to each joint motor of the main robotic arm.

[0080] Based on the dynamic model of the robotic arm, the gravitational torque that each joint needs to withstand is calculated using the current joint angle.

[0081] As an example, this can be achieved using the Newton-Euler iterative algorithm or a pre-computed lookup table.

[0082] Specifically, after the parameters of each link of the main robotic arm are accurately modeled, the gravitational torque under any posture is calculated using the Newton-Euler method: in, T To control the cycle, q For joints, m denoted as , where is the weight of the main robotic arm and g is the acceleration due to gravity.

[0083] Next, the gravitational torque It is converted into the gravity compensation current required by the motors of each joint, so as to achieve active gravity cancellation throughout the entire stroke range.

[0084] In this way, by actively outputting a torque to counteract gravity, the operator feels that the robotic arm is in a zero-gravity state, which allows for easy movement and positioning, and more accurate and stable acquisition of joint angle data from the main robotic arm, which is isomorphic to the robotic arm.

[0085] In some embodiments, the method further includes: The dynamic model is optimized, and the gravitational torque of the main robotic arm in the current posture is determined based on the optimized dynamic model; The dynamic model is optimized through parameter identification, which is based on the motion data and current data of the main robotic arm under various excitation trajectories.

[0086] The effectiveness of gravity compensation depends on the accuracy of the dynamic model, so it is necessary to optimize the model parameters through methods such as parameter identification.

[0087] In this embodiment, by identifying parameters under multiple operating conditions, the accuracy of the dynamic model is optimized by correcting friction, transmission clearance, and inertial coupling terms. Ultimately, this allows the main boom to hover stably in any posture, and the operator can release the handle at any time, greatly improving the comfort of human-machine interaction.

[0088] Through the above embodiments, the main robotic arm integrates a high-precision gravity compensation function and, based on a dynamic model optimized by parameter identification, achieves hands-free hovering in any posture throughout the entire stroke. This allows the operator to release the main robotic arm at any time without affecting system stability, significantly reducing physical fatigue and operational pressure during long-term demonstration data acquisition, and improving the comfort and sustainability of long-term data acquisition tasks.

[0089] Step two: Switching between operating modes and handing over control.

[0090] In some embodiments, the current operation phase includes a second operation phase in which the robotic arm and the work object are in contact or nearly in contact. The step of switching the control mode of the robotic arm to the target control mode in response to a mode switching trigger command includes: In response to the mode switching trigger command, the control mode of the robotic arm is switched to the second control mode.

[0091] The mode switching trigger command is generated based on the current operation stage of the robotic arm, which represents different operational requirements for the robotic arm. For example, when the operation enters a fine-tuning stage (such as about to contact the fabric, perform layering, or fine-tune the clamping position), the system initiates the control mode switching mechanism through the mode switching trigger command.

[0092] In this embodiment, there are two ways to generate the mode switching trigger command: one is manual triggering, and the other is automatic triggering.

[0093] Automatic triggering refers to the system continuously monitoring some parameters of the robotic arm. When the parameters reflect the current operation stage, the system automatically selects and switches the target control mode according to the current operation stage.

[0094] As an example, the mode switching trigger command is generated in the following ways: Obtain the relative position information and / or contact force information between the robotic arm and the work object; When the relative position information and / or contact force information satisfy the first condition characterizing that the current operation stage of the robotic arm is in the second operation stage, a mode switching trigger command to switch to the second control mode is generated. The first condition includes: the relative position information is less than the safe position threshold, and / or the contact force information is greater than zero.

[0095] The system continuously monitors the distance between the end of the robotic arm and the fabric. When this distance is less than the safe position threshold (e.g., 5cm), it indicates that the end of the robotic arm is in contact with or close to contacting the work object.

[0096] Similarly, the system can continuously monitor the relative speed between the end of the robotic arm and the fabric. When the relative speed is lower than the set value, it indicates that the end of the robotic arm is in contact with or close to contacting the work object.

[0097] For example, by continuously monitoring the contact force between the end of the robotic arm and the fabric, the force sensor at the end of the arm determines that the contact force is greater than zero, indicating that the end of the robotic arm is in contact with or close to contact with the work object.

[0098] In this state, the system automatically generates a mode switching trigger command to switch to the second control mode.

[0099] For manual triggering, it means that the operator observes the current working stage of the robotic arm and then automatically selects and switches the target control mode according to the current working stage.

[0100] As an example, the mode switching trigger command is generated in the following ways: In response to a mode switching trigger request input by the user based on the current operation stage of the robotic arm, a corresponding mode switching trigger instruction is generated based on the mode switching trigger request.

[0101] When the operator determines that a control mode needs to be switched, they can initiate the control mode switching mechanism through an external trigger signal.

[0102] For example, a mode switching trigger command can be generated by manually operating a dedicated physical button on the device, such as a switch button.

[0103] In some cases, to ensure the accuracy of the mode switching trigger command, a series of processes are performed on the command, including debouncing the key signal to prevent false triggers caused by mechanical contact bounce. Alternatively, a valid trigger may only be confirmed if the key is pressed and held for a certain period of time to avoid accidental touches. Or, the trigger command may be latched by the system and only cleared after the switching process has been fully executed to prevent duplicate triggers.

[0104] Subsequently, in response to the mode switching trigger command, the control mode of the slave robotic arm is switched to the target control mode, and the movement of the slave robotic arm is controlled based on the target control mode.

[0105] To ensure a seamless transfer of control during the handover process and avoid conflicts from multiple sources of commands, the following security policies are implemented during the handover process: Strategy 1 involves clearing instructions and buffers to prevent residual data from causing malfunctions.

[0106] As an example, before switching the control mode from the robotic arm to the target control mode, the method further includes: Stop instruction transmission corresponding to the current control mode and clear any unexecuted instructions from the instruction buffer.

[0107] Stop the data transmission logic of master-slave joint mapping, pause the joint command transmission from master robot arm to slave robot arm, and mark generated but unsent trajectory data as invalid and clear it from the transmission buffer.

[0108] Strategy 2: Ensure the orderly switching of operations to avoid command conflicts.

[0109] As an example, the operator actively presses the "End Joint Mapping" button on the main arm control panel or a dedicated button. This sends a signal to the master controller that the operator has relinquished direct control of the main arm and is ready to exit the current mode. The operator then presses the "Start Handle Control" button on the handle. This sends a signal to the master controller that the operator requests to activate the fine control mode. After receiving these two signals in sequence, the master controller initiates a switch from the first control mode to the second control mode. This means that, based on the principle that mode switching is achieved solely through the mode switching trigger command corresponding to the target control mode, to ensure command accuracy and avoid command conflicts, the mode switching trigger command will be responded to according to the set switching order. For example, if the master controller only receives the handle control command, that command will be ignored.

[0110] Similarly, when switching from the second control mode to the first control mode, the mode switching trigger command is responded to according to the set switching order. For example, you must first press "End Handle Control" on the handle, and then press "Start Joint Mapping" on the main robotic arm side.

[0111] It should be noted that, in addition to setting physical buttons on robotic arms or manually operated devices, other methods such as voice control and virtual buttons can also be used to generate mode switching trigger commands, which are not specifically limited here.

[0112] As an example, the orderliness of switching operations can be achieved by setting an instruction source mutex lock.

[0113] Specifically, the system maintains a global flag indicating the current valid instruction source, ensuring there are no valid instruction sources before switching. During the switching process, it continuously monitors for valid control modes; when no valid control mode is detected, it indicates the previous control mode has ended / exited. Simultaneously, all control modes must check this flag before generating a mode switching trigger command; only currently valid sources can be written to the instruction buffer. Therefore, a control mode switch is executed in response to the mode switching trigger signal. In this embodiment, the flag changes from the joint mapping mode to the manual operation device input mode.

[0114] Strategy 2: Smooth transition of target pose to avoid sudden jumps at the end.

[0115] As an example, switching the control mode from the robotic arm to the target control mode includes: Based on the actual end-effector pose at the moment of switching from the robotic arm, a motion trajectory is generated that smoothly transitions to the initial desired end-effector pose in the target control mode. Control commands are generated based on the motion trajectory to control the robotic arm to move along the motion trajectory.

[0116] The actual end-effector pose of the slave robot arm is obtained when the master-slave joint mapping stops or at the moment of switching. This pose is used as a switching reference point. Combined with the initial desired end-effector pose of the slave robot arm at the moment when the manual operation device takes over, a smooth transition motion trajectory is generated. Based on this motion trajectory, the robot arm moves smoothly to the initial desired end-effector pose, avoiding inertial shocks caused by sudden stops or switching, and ensuring seamless spatiotemporal connection between the two control modes.

[0117] The aforementioned smooth transition process is achieved by performing a smooth Cartesian straight line or circular arc interpolation between the moment the master-slave joint mapping stops or the instant of switching and the first effective control cycle calculated by the manual operation equipment control module.

[0118] In some cases, the smooth transition of the motion trajectory is also driven in real time by reverse mapping, so that the operator can feel the main robotic arm being smoothly moved to the new starting point, ensuring a seamless switch in vision and touch.

[0119] After switching control modes, the target control mode is executed to control the movement of the robotic arm.

[0120] In some embodiments, when the target control mode is the second control mode, the method further includes controlling the movement of the robotic arm based on the target control mode, wherein the robotic arm is moved according to the target control mode. Activate the signal input channel corresponding to the second control mode; The incremental pose signal received through the signal input channel for the robotic arm end effector is superimposed with the current actual pose of the end effector to obtain the desired end effector pose. Based on the desired end pose, the target value of the joint space is obtained by inverse kinematics solution; The target value of the joint space is sent to the servo system of the slave robot to control the slave robot to perform end effector incremental motion.

[0121] In this embodiment, the manually operated device is a device that supports six degrees of freedom input, such as one with dual joysticks, trigger buttons, gyroscopes, etc.

[0122] For example, a manually operated device uses a control handle. The movement of the left joystick controls the movement of the robotic arm. x The movement on the axis is controlled by the forward and backward movement of the left joystick from the end of the robotic arm. y Movement on the axis is controlled by the right joystick, rotating from the end of the robotic arm. z shaft or y Rotation of axes, etc., are not specifically limited here.

[0123] First, activate the handle input channel to begin receiving incremental pose signals from the robotic arm's end effector input by the operator via the six-DOF handle. Including three-dimensional translation increments and three-dimensional rotation increment .

[0124] In some examples, after activating the signal input channel corresponding to the second control mode, the method further includes: Reliability processing is performed on the incremental pose signal received through the signal input channel for the robotic arm end effector. The processed incremental pose signal is superimposed on the actual pose of the current end effector to obtain the desired end effector pose.

[0125] Reliability processing includes, but is not limited to, zero-point calibration and dead-zone filtering. Zero-point calibration or dead-zone filtering is performed on all input signals to prevent drift interference.

[0126] For example, after a switching command is triggered, the system continuously collects multiple handle readings, calculates the average value of each channel as the zero-point reference value for that channel, and subtracts the corresponding zero-point reference value from all subsequent readings. Zero-point calibration compensates for the DC bias of the handle sensors and circuitry, keeping the robotic arm stationary when there is no input.

[0127] For example, a fixed dead-zone threshold is set for each input channel to filter out minor, involuntary inputs. Dead-zone filtering effectively eliminates unintended inputs such as inaccurate joystick return and slight tremors in the operator's hand, preventing uncontrolled movements of the robotic arm.

[0128] Step 3: Handle control and inverse kinematics solution based on pose increment commands using the end-tool coordinate system (TCP, Tool Center Point).

[0129] In handle control mode, the system no longer relies on the main arm joint mapping, but controls the slave arm based on the incremental pose signal output by the handle.

[0130] The slave controller obtains the actual pose of the current end effector of the robotic arm by reading the encoders of each joint in real time. It also acquires the incremental pose signal received through the signal input channel. Then, the incremental pose signal, after reliability processing, is superimposed onto the actual pose from the current end effector of the robotic arm. The new desired end pose is obtained: .in, This represents quaternion multiplication.

[0131] Next, the analytical or numerical inverse kinematics solver (IK Solver) is invoked to determine the desired end-effector pose. Calculate the corresponding seven-axis joint space target value The solution process incorporates joint constraints, singularity avoidance, and minimum change criteria to ensure the feasibility and smoothness of the solution.

[0132] Final joint space target value As a new instruction issued to the robotic arm servo system, it enables precise fine-tuning of the end-effector TCP, making it particularly suitable for high-precision tasks such as fabric edge alignment and single-layer peeling.

[0133] Step 4: Synchronize the master-to-master reverse mapping with the master arm state.

[0134] Regardless of the system's control mode, the system maintains the operation of the master-slave reverse mapping channel. That is, a reverse state feedback channel is established from the slave end to the master end, synchronizing the actual joint positions and force information of the robotic arm to the master robotic arm. This reverse state feedback channel operates in parallel with the forward control channel, ensuring that the bidirectional data streams do not interfere with each other. Based on this, the system encapsulates the actual state of the robotic arm (joint angles, speeds, torques) into data packets and sends them to the master end at a set frequency. This set frequency can be the same as the data acquisition frequency from the master end to the slave end.

[0135] In some embodiments, the method further includes: Obtain the real-time joint state information of the robotic arm; The real-time joint state information is mapped to the isomorphic main robotic arm, and the main robotic arm is driven to perform corresponding posture synchronization.

[0136] In the reverse mapping of the master-slave joint mapping mode, the slave controller collects the encoder readings of each joint in real time and calculates the real-time joint state information of the slave robotic arm. Using motion range parameters from each joint of the robotic arm, real-time joint status information is obtained. Normalization is performed, and the normalized joint state information, along with a timestamp and data validity flag, is encapsulated and transmitted back to the main arm via a high-speed communication link to drive it to synchronously update its posture. For example, based on the current pose of the main robotic arm and real-time joint state information... Obtain the joint angle change after reverse mapping. The main robotic arm moves according to the changes in joint angles, driving it to update its posture synchronously.

[0137] In this mode, the operator actively moves the main robotic arm, but the main robotic arm is simultaneously affected by the reverse mapping. When the secondary robotic arm cannot reach the commanded position due to environmental constraints, the reverse mapping pulls the main robotic arm back, allowing the operator to feel resistance and providing a natural force feedback.

[0138] In reverse mapping under controller mode, although forward control has switched to controller incremental mode, the reverse mapping channel remains active, continuously feeding back feedback from the robotic arm joint status to the main robotic arm. The main robotic arm is no longer directly dragged by the operator, but is entirely driven by reverse mapping, causing the main robotic arm to passively follow the movement of the slave robotic arm. This allows the operator to perceive the actual status of the slave arm through both vision and touch, improving control intuition and task success rate.

[0139] In this embodiment, the reverse mapping also employs the same interpolation and smoothing algorithms as the forward mapping to ensure bidirectional transparency. This mechanism enhances the immersive experience of the system, providing the operator with a richer perception of the remote environment, which helps the operator determine the current contact state and facilitates precise fabric application.

[0140] Step 5: Force sensing fusion and admittance control closed loop.

[0141] In some embodiments, the method further includes: Obtain the contact force information between the end effector of the robotic arm and the work object; Based on the contact force information, the end-effector motion compensation amount is obtained through the admittance control model; The end-effector motion compensation amount is used to correct the desired end-effector pose generated based on the pose increment signal; The movement of the robotic arm is controlled based on the corrected desired pose.

[0142] In this embodiment, a compact six-dimensional force or torque sensor is installed between the end flange of the robotic arm and the gripper to collect contact force information between the end of the robotic arm and the fabric in real time. .

[0143] In some cases, the raw data is processed by standard Kalman filtering to suppress noise and drift, resulting in a stable and reliable force estimate. In sewing scenarios, fabric contact forces typically involve complex low-frequency oscillations caused by fabric deformation, rebound, and slippage. The model prediction capabilities of Kalman filters help smooth these oscillations and extract stable contact trends.

[0144] Subsequently, the processed contact force information is input into the admittance control model. Admittance control is an outer-loop control strategy with force input and position / velocity output. A second-order admittance model is used: in, M Let B be the mass and B be the damping. K For stiffness, , , These are the position deviation, speed compensation, and acceleration compensation of the end effector, respectively. The parameters were set through experimental calibration, and the optimal settings were selected for scenarios involving light touches on the fabric: ).

[0145] The output of the admittance control model is the compensation velocity that the terminal should produce. For compensation speed Integrating, we obtain the end-effector compensation. The end-effector motion compensation is superimposed on the current desired end-effector pose, thereby forming an active compliant response when controlling the movement of the robotic arm based on the corrected desired pose.

[0146] In this embodiment, the admittance parameter is optimized for flexible materials to ensure the stability and safety of touch interaction.

[0147] In some embodiments, after acquiring the contact force information between the robotic arm end effector and the work object, the method further includes: When the contact force corresponding to the contact force information exceeds a preset safety threshold, the generation of the desired end pose based on the pose increment signal is paused. The robotic arm is controlled to execute a retreat trajectory generated based on the admittance control model.

[0148] Under normal circumstances, the admittance model provides compensation to fine-tune the pose and achieve compliance; under abnormal circumstances, it automatically pauses command response and activates the protection mechanism to effectively prevent the fabric from being pulled, deformed or torn.

[0149] Specifically, in handle control mode, once the contact force is detected to exceed a preset safety threshold (e.g., ... The system immediately activates the force protection mechanism: First, pause the controller command response.

[0150] Force pause the controller incremental command processing thread, preventing any new incremental command ΔT from being generated.

[0151] Next, admittance control ensures the force value decreases.

[0152] The system executes a linear retreat trajectory generated by admittance control, starting from the current position and proceeding along the retreat direction until the force value falls back to within the safe threshold range, ensuring that the fabric is not damaged.

[0153] Finally, during the retreat process, the contact force is continuously monitored. When the force value drops below the safety threshold, the emergency retreat movement stops, and the robotic arm remains in the new position. Handlebar command response is then gradually restored. In some cases, handlebar command response is restored only after the force value has dropped below the safety threshold and remained stable for a certain period.

[0154] In other examples, a two-tiered force protection mechanism can be implemented. Specifically, two safety thresholds are set. The first safety threshold serves as a warning threshold, triggering a visual or auditory warning to alert the operator that excessive force has been applied or is about to be applied. The second safety threshold has a higher safety limit than the first threshold; once the second safety threshold is exceeded, a mandatory force protection mechanism is immediately triggered. This protection mechanism can provide warnings and corrections before the contact force exceeds the limit, while also preventing malfunctions caused by sensor noise or other issues.

[0155] To enhance operator awareness, the system provides appropriate feedback and stops accumulating when the contact force reaches its limit. Feedback methods may include, but are not limited to, handle vibration, interface flashing, and audio cues.

[0156] Through the admittance control in this embodiment, rapid force relief and active yielding are used to strictly limit the contact force within the range that the fabric material can withstand, thereby achieving a high success rate of single-layer fabric separation.

[0157] After completing the fine operation on the work object according to the second control mode, it is necessary to return the robotic arm to the initial position or move it to other set positions. At this time, it is necessary to move from the end of the robotic arm within a large range. Select the first control mode to control the movement from the robotic arm.

[0158] Specifically, the current operation phase includes a first operation phase in which the robotic arm and the work object are in a non-contact state. The step of switching the control mode of the robotic arm to the target control mode in response to a mode switching trigger command includes: In response to the mode switching trigger command, the control mode of the robotic arm is switched to the first control mode.

[0159] In the first operational phase, where the robotic arm and the work object are in a non-contact state, this phase represents a requirement for a wide range of movements of the robotic arm. At this stage, the robotic arm needs to be controlled via master-slave joint mapping, thus generating a mode-switching trigger command to switch the control mode to the first control mode. The control method for master-slave joint mapping is the same as in step one described above, and will not be repeated here.

[0160] It should be noted that when switching from the second control mode to the first control mode, it is necessary to ensure that the state of the slave robot arm and the master robot arm are consistent. This can be achieved in the following two ways: Method 1: When the handle controls the movement of the robotic arm, the process of reverse mapping from the master to the slave arm and synchronizing the master arm state in step four is used to maintain the consistency between the slave and master robotic arms during mode switching.

[0161] Method 2: When the handle control ends, a smooth synchronous trajectory is planned based on the current pose of the slave robot and the current pose of the master robot, and the master robot is controlled to move according to the synchronous trajectory until it is consistent with the state of the slave robot.

[0162] Through the above embodiments, a dual-mode adaptive collaborative control teleoperation architecture is set up to coordinate the operation and switching of the two control modes, balancing efficiency and accuracy during operation. Specifically, in the large-scale movement phase (i.e., the first operation phase), a full-dimensional joint space mapping mode from the master robotic arm to the slave robotic arm (i.e., the first control mode) is adopted to achieve fast, intuitive, and low-latency motion reproduction, significantly improving high-speed response capability and operation efficiency. In the fine operation phase (i.e., the second operation phase), the system switches to a TCP incremental control mode driven by manual operation equipment (i.e., the second control mode), combined with real-time inverse kinematics solving, to achieve micron-level precise control of the end effector, providing high-precision and compliant control capabilities when near or in contact with the work object, suitable for high-difficulty operations such as fabric layering and edge alignment. This system architecture and teleoperation method significantly improve the success rate and stability of single-layer fabric separation.

[0163] Furthermore, the system also provides a universal data acquisition and demonstration learning platform for complex and dexterous operation tasks in intelligent manufacturing, and has broad application value.

[0164] In some embodiments, the method further includes: During the control of the slave robotic arm, a dataset for training the slave robotic arm to autonomously execute tasks is recorded simultaneously; The dataset includes at least the joint motion sequence, end-effector pose sequence, and contact force information sequence of the robotic arm, and the sequences are aligned by timestamps.

[0165] Throughout the entire remote operation control process, the master controller synchronously collects multi-source data, including but not limited to the joint motion sequence formed by reading joint angle data in real time from the joint encoder of the robotic arm, the end pose sequence formed by calculating the end pose in real time based on the joint angle from the kinematic model of the robotic arm, and the contact force information sequence formed by reading contact force and torque information in real time from the six-dimensional force sensor at the end of the robotic arm.

[0166] In some examples, each data set is stamped with a precise timestamp at the same moment. The data sequences are strictly aligned on the timeline using the timestamps, ensuring that the joint state, end-effector pose, and contact force information at the same moment correspond accurately.

[0167] In some cases, the recorded datasets are organized by task or time segment and stored in non-volatile memory.

[0168] This process records data throughout the entire teleoperation task. Because human operators can demonstrate complex and dexterous movements during the human-robot collaboration phase of teleoperation, reliable motion and force data are provided for robot learning. Therefore, the recorded dataset can be transformed into structured, quantifiable machine learning data for training robot imitation learning / reinforcement learning algorithms, laying the data foundation for the subsequent autonomy and intelligence of the robotic arm.

[0169] In this embodiment, a set of efficient, precise, compliant, safe, and low-load isomorphic teleoperation solutions are constructed by adopting a dual-mode collaborative control strategy, mode switching logic, force-sensing safety mechanism, reverse mapping synchronization, and main arm gravity compensation function. In fine operation scenarios of flexible materials such as single-layer fabric separation, it not only significantly improves the success rate and stability of single-layer fabric separation, but also provides a high-quality human demonstration data foundation for robot skill learning, which has important theoretical significance and broad industrial application prospects.

[0170] This application provides a remote operation control method, device, and system, which have the following beneficial effects: Two control modes are implemented: a first control mode and a second control mode. In the first control mode, based on joint motion signals from the isomorphic master robotic arm, the slave robotic arm is controlled to perform joint space mapping motion. This isomorphic mapping enables the slave robotic arm to perform tasks such as rapid positioning or movement over a wide range. In the second control mode, based on incremental pose signals from the end effector of the slave robotic arm, the slave robotic arm is controlled to perform incremental end effector motion. This enables precise and compliant operation during the near-contact phase between the end effector and the work object. By seamlessly and safely switching between the joint space isomorphic mapping and incremental end effector control modes, the system can balance the speed and accuracy requirements of teleoperation. Simultaneously, a corresponding mode switching trigger command is generated when the current work stage of the slave robotic arm is determined. This command switches the control mode of the slave robotic arm to the mode corresponding to the current work stage, thus meeting the operational requirements of the slave robotic arm at that stage. The command switching ensures the smoothness of the operation process. In summary, the adaptive dual-mode control based on work stage perception meets the needs of different work scenarios and provides high-quality demonstration data for the robot's subsequent autonomous learning and intelligent decision-making capabilities.

[0171] Figure 3 An example is a schematic diagram of the physical structure of a remote control device, such as... Figure 3 As shown, the remote control device may include a processor 810, a communication interface 820, a memory 830, and a communication bus 840. The processor 810, communication interface 820, and memory 830 communicate with each other via the communication bus 840. The processor 810 can call logical instructions from the memory 830 to execute remote control methods.

[0172] Furthermore, the logical instructions in the aforementioned memory 830 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0173] On the other hand, this application also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer is able to perform the teleoperation control methods provided by the above methods.

[0174] In another aspect, this application also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to perform the teleoperation control methods provided by the methods described above.

[0175] It should be noted that the technical solutions or features described in the above embodiments can be combined or supplemented with each other without conflict. The scope of protection of this application is not limited to the precise structures described in the above embodiments and shown in the accompanying drawings; all modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A remote operation control method, characterized in that, The method includes: Receive a mode switching trigger instruction, which is generated based on the current operation stage of the robotic arm, and the current operation stage is used to characterize different operational requirements for the robotic arm; In response to a mode switching trigger command, the control mode of the slave robotic arm is switched to the target control mode, and the movement of the slave robotic arm is controlled based on the target control mode; The target control mode is either a first control mode or a second control mode. In the first control mode, the slave robot arm is controlled to perform joint space mapping motion based on the joint motion signal from the isomorphic master robot arm. In the second control mode, the slave robot arm is controlled to perform end-effector incremental motion based on the incremental pose signal for the end effector of the slave robot arm.

2. The teleoperation control method as described in claim 1, characterized in that, The current operation phase includes a first operation phase in which the robotic arm and the work object are in a non-contact state. The step of switching the control mode of the robotic arm to the target control mode in response to a mode switching trigger command includes: In response to the mode switching trigger command, the control mode of the robotic arm is switched to the first control mode.

3. The teleoperation control method as described in claim 1, characterized in that, The current operation phase includes a second operation phase in which the robotic arm is in contact with or near contact with the work object. The step of switching the control mode of the robotic arm to the target control mode in response to a mode switching trigger command includes: In response to the mode switching trigger command, the control mode of the robotic arm is switched to the second control mode.

4. The teleoperation control method as described in claim 3, characterized in that, The generation methods for the mode switching trigger command include: Obtain the relative position information and / or contact force information between the robotic arm and the work object; When the relative position information and / or contact force information satisfy the first condition characterizing that the current operation stage of the robotic arm is in the second operation stage, a mode switching trigger command to switch to the second control mode is generated. The first condition includes: the relative position information is less than the safe position threshold, and / or the contact force information is greater than zero.

5. The teleoperation control method as described in claim 1, characterized in that, The generation methods for the mode switching trigger command include: In response to a mode switching trigger request input by the user based on the current operation stage of the robotic arm, a corresponding mode switching trigger instruction is generated based on the mode switching trigger request.

6. The teleoperation control method as described in claim 1, characterized in that, Switching the control mode from the robotic arm to the target control mode includes: Based on the actual end-effector pose at the moment of switching from the robotic arm, a motion trajectory is generated that smoothly transitions to the initial desired end-effector pose in the target control mode. Control commands are generated based on the motion trajectory to control the robotic arm to move along the motion trajectory.

7. The teleoperation control method as described in claim 1, characterized in that, Before switching the control mode from the robotic arm to the target control mode, the method further includes: Stop instruction transmission corresponding to the current control mode and clear any unexecuted instructions from the instruction buffer.

8. The teleoperation control method as described in claim 1, characterized in that, When the target control mode is the second control mode, the method further includes controlling the movement of the robotic arm based on the target control mode: Activate the signal input channel corresponding to the second control mode; The incremental pose signal received through the signal input channel for the robotic arm end effector is superimposed with the current actual pose of the end effector to obtain the desired end effector pose. Based on the desired end pose, the target value of the joint space is obtained by inverse kinematics solution; The target value of the joint space is sent to the servo system of the slave robot to control the slave robot to perform end effector incremental motion.

9. The teleoperation control method as described in claim 8, characterized in that, After activating the signal input channel corresponding to the second control mode, the method further includes: Reliability processing is performed on the incremental pose signal received through the signal input channel for the robotic arm end effector. The processed incremental pose signal is superimposed on the actual pose of the current end effector to obtain the desired end effector pose.

10. The teleoperation control method as described in claim 9, characterized in that, The method further includes: Obtain the contact force information between the end effector of the robotic arm and the work object; Based on the contact force information, the end-effector motion compensation amount is obtained through the admittance control model; The end-effector motion compensation amount is used to correct the desired end-effector pose generated based on the pose increment signal; The movement of the robotic arm is controlled based on the corrected desired pose.

11. The teleoperation control method as described in claim 10, characterized in that, After obtaining the contact force information between the robotic arm end effector and the work object, the method further includes: When the contact force corresponding to the contact force information exceeds a preset safety threshold, the generation of the desired end pose based on the pose increment signal is paused. The robotic arm is controlled to execute a retreat trajectory generated based on the admittance control model.

12. The teleoperation control method as described in claim 1, characterized in that, When the target control mode is the first control mode, the step of controlling the movement of the robotic arm based on the target control mode includes: Obtain the joint angle data of the master robotic arm, which is isomorphic to the slave robotic arm; A smooth trajectory sequence is generated based on the joint angle data, the smooth trajectory sequence including multiple target joint angles arranged in chronological order; The smooth trajectory sequence is sent to the servo system of the slave robot to control the slave robot to perform motion tracking according to the target joint angle.

13. The teleoperation control method as described in claim 12, characterized in that, Before acquiring the joint angle data of the master robotic arm, which is isomorphic to the slave robotic arm, the method further includes: A gravity-compensating current is applied to the main robotic arm to bring it into a state of force balance.

14. The teleoperation control method as described in claim 13, characterized in that, Applying a gravity-compensating current to the main robotic arm includes: Based on the isomorphic dynamic model of the main manipulator, the gravitational torque of the main manipulator in the current posture is determined; The gravitational torque is converted into gravity compensation current for the motors of each joint of the main robotic arm; The gravity compensation current is applied to each joint motor of the main robotic arm.

15. The teleoperation control method as described in claim 14, characterized in that, The method further includes: The dynamic model is optimized, and the gravitational torque of the main robotic arm in the current posture is determined based on the optimized dynamic model; The dynamic model is optimized through parameter identification, which is based on the motion data and current data of the main robotic arm under various excitation trajectories.

16. The teleoperation control method as described in claim 1, characterized in that, The method further includes: Obtain the real-time joint state information of the robotic arm; The real-time joint state information is mapped to the isomorphic main robotic arm, and the main robotic arm is driven to perform corresponding posture synchronization.

17. The teleoperation control method as described in claim 1, characterized in that, The method further includes: During the control of the slave robotic arm, a dataset for training the slave robotic arm to autonomously execute tasks is recorded simultaneously; The dataset includes at least the joint motion sequence, end-effector pose sequence, and contact force information sequence of the robotic arm, and the sequences are aligned by timestamps.

18. An electronic device, characterized in that, The device includes a memory, a processor, and a teleoperation control program stored in the memory and executable on the processor, wherein the processor, when executing the teleoperation control program, implements the teleoperation control method as described in any one of claims 1 to 17.

19. A remote operation control system, characterized in that, The system includes a homogeneous master robotic arm and a slave robotic arm, a slave controller for tracking the motion of the slave robotic arm, and an electronic device as described in claim 18.

20. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a teleoperation control program, which, when executed, implements the steps of the teleoperation control method as described in any one of claims 1 to 17.

Images