Parallel gripper device based on visual servoing and force feedback and control method
Patent Information
- Application Number
- CN202611276205.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-21
- Publication Date
- 2026-09-29
AI Technical Summary
这种“先看后动”的模式无法在夹持过程中根据视觉反馈实时纠正因物体形变、夹爪偏摆带来的微米级定位误差，更无法实现力-视觉融合的自适应夹取策略
1.本发明所述的一种基于视觉伺服与力反馈的平行夹钳装置，通过将视觉伺服与力反馈技术深度融合，实现了夹钳对复杂、精密任务的高适应性操作；夹钳组件中平行设置的移动夹爪与固定夹爪确保了夹持过程中物体与夹爪保持面接触，避免了应力集中，有效保护易碎物体免受损坏；末端力感知模块直接测量夹持界面的真实力值与压力分布，实现了精确的闭环力控制；视觉模块与力感知模块的协同工作，使夹钳能够实时感知并响应操作过程中的动态变化，显著提升了操作安全性与任务成功率。
Smart Images
Figure CN122829794A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of precision operation and automated control technology, and in particular to a parallel clamp device and control method based on vision servo and force feedback. Background Technology
[0002] In fields such as precision electronic assembly, handling of fragile components, and automation in life sciences, mechanical clamps need to perform highly reliable gripping, handling, and assembly operations. These tasks typically require the clamps to keep the jaws parallel throughout their movement to prevent objects from tilting, slipping, or experiencing uneven force; they also require precise closed-loop control of the clamping force to prevent damage to fragile workpieces; in addition, they need to rely on vision systems for precise positioning to compensate for positioning errors.
[0003] However, existing technologies cannot simultaneously meet the above requirements: Firstly, regarding maintaining the parallelism of the grippers, common solutions include using external linear guides or scissor-type linkage mechanisms. The former is bulky, complex in structure, and expensive; the latter is prone to elastic deformation under load, leading to loss of parallelism and making it difficult to maintain accuracy under high-speed motion or variable load conditions.
[0004] Secondly, regarding clamping force control, most industrial clamps employ simple open-loop pneumatic or motor position control, making it impossible to sense and adjust the clamping force in real time, which easily leads to insufficient clamping force or overload. Even in the few solutions equipped with force sensors, the sensors often cannot directly measure the actual force value at the end-grip interface.
[0005] Furthermore, in terms of visual assistance, existing systems typically separate visual positioning from gripping operations: after the vision system completes a coarse positioning, the gripper performs a fixed action according to the program. This "see first, then move" mode cannot correct micron-level positioning errors caused by object deformation or gripper misalignment in real time based on visual feedback during the gripping process, and it cannot achieve an adaptive gripping strategy that integrates force and vision. Summary of the Invention
[0006] Therefore, the technical problem to be solved by the present invention is to overcome the shortcomings of the prior art and provide a parallel clamping and control method based on visual servoing and force feedback, which can deeply integrate a high-rigidity active parallel gripper mechanism, closed-loop control based on end-effector direct force sensing and real-time visual servo positioning, fundamentally improving the clamp's adaptability to complex and precision tasks and operational safety.
[0007] To address the aforementioned technical problems, this invention provides a parallel clamping device based on visual servoing and force feedback, comprising: Three-axis motion platform; A clamp drive assembly is mounted on the three-axis motion platform; A clamp assembly is connected to the clamp drive assembly. The clamp assembly includes a fixed clamp and a movable clamp. The fixed clamp is provided with a fixed jaw, and the movable clamp is provided with a movable jaw. The movable jaw is arranged parallel to the fixed jaw. An end force sensing module is installed on the fixed gripper and / or the movable gripper to measure the gripping force and pressure distribution in real time. The vision module is used to acquire the recognition information and pose information of the target object; The controller, which is connected to the vision module and the end force sensing module respectively, is configured to plan an initial gripping strategy based on the vision information and control the gripping force according to the end force sensing information.
[0008] In one embodiment of the present invention, the three-axis motion platform includes an X-axis motor, a Y-axis motor and a Z-axis motor. The Y-axis motor is mounted on the X-axis motor moving rail, and the Z-axis motor is mounted on the Y-axis motor moving rail, providing translational degrees of freedom of the X, Y and Z axes for the clamp drive assembly.
[0009] In one embodiment of the present invention, the clamp drive assembly is disposed on the Z-axis motor moving rail and includes a clamp motor, a clamp motor fixed rail, a clamp motor moving rail, a fixed clamp mounting block and a movable clamp mounting block. The fixed clamp mounting block is fixedly connected to the clamp motor fixed rail, and the movable clamp mounting block is fixedly connected to the clamp motor moving rail.
[0010] In one embodiment of the present invention, the fixed clamp mounting block is fixedly connected to the clamp motor fixed rail by fastener one; the movable clamp mounting block is fixedly connected to the clamp motor movable rail by fastener one; the fixed clamp is fixedly connected to the fixed clamp mounting block by positioning element two and fastener two, and the movable clamp is fixedly connected to the movable clamp mounting block by positioning element one and fastener two.
[0011] In one embodiment of the present invention, the fixing clamp includes a fixing clamp one, a positioning block and a fixing clamp two; the fixing clamp one and the fixing clamp two are fixedly connected to the fixing clamp mounting block after being positioned by the positioning member two; the fixing claw is disposed on the fixing clamp two; the positioning block is disposed on the fixing clamp one and is used to pre-position the target object.
[0012] In one embodiment of the present invention, the end force sensing module includes a flexible pressure sensing film disposed at one end of the fixed gripper near the movable gripper.
[0013] In one embodiment of the present invention, the identification information includes the mass and fragility level of the target object; the pose information includes the three-dimensional pose of the target object.
[0014] In one embodiment of the present invention, the controller is configured to: switch control modes according to the end force sensing information during the gripping process, and finely adjust the gripping posture and force by fusing real-time visual information and end force information to cope with dynamic changes in slippage and off-center loading.
[0015] In one embodiment of the present invention, the collaborative fine-tuning strategy of the controller includes a pressure distribution equalization algorithm: when the end force sensing module detects that the force difference between the two sides of the clamp exceeds the threshold ΔF, and the vision module simultaneously detects that the object tilts beyond the threshold Δθ, the controller calculates a small corrective translation δx and drives the clamp to perform the translation while keeping the total force unchanged, so that the force difference between the two sides returns to the safe range.
[0016] Secondly, in order to solve the above-mentioned technical problems, the present invention provides a clamping control method based on the parallel clamping device described in the first aspect, comprising the following steps: The vision module acquires the recognition and pose information of the target object. Based on the recognition and pose information of the target object obtained by the vision module, an initial gripping strategy is planned and the gripper is controlled to move to the pre-contact point. After moving to the pre-contact point based on the visual servo trajectory, the target is initially fixed by pressing down the positioning block in the fixed clamp. Then, the force control mode is switched, and the moving clamp is slowly closed according to the real-time feedback of the end force sensing module. When the contact force is detected, the actual contact point is recorded and compared with the visually estimated point to complete the online calibration of visual error. In force control mode, the clamping force is steadily increased to the target value, and the object's pose is monitored simultaneously based on visual signals. When uneven pressure distribution and object tilt are detected, visual information and force information are fused together, and a small pose translation or rotation is performed while keeping the total force constant, so as to make the pressure distribution uniform. During the handling process, the clamping force is kept stable in a closed loop and slippage is detected simultaneously based on force sensing signals, and the object's pose is monitored based on visual signals. When high-frequency vibration or force change is detected and the object's feature points move, slippage is determined to have occurred, and the clamping force is increased or the posture is adjusted to re-stabilize the object. When an abnormality is detected, the force or pose of the gripper is adjusted by fusing visual and force information.
[0017] Compared with the prior art, the above-described technical solution of the present invention has the following advantages: 1. The parallel clamping device based on visual servoing and force feedback described in this invention achieves highly adaptable operation of the clamp for complex and precise tasks by deeply integrating visual servoing and force feedback technologies. The parallel moving jaws and fixed jaws in the clamping assembly ensure that the object and the jaws maintain surface contact during clamping, avoiding stress concentration and effectively protecting fragile objects from damage. The end effector force sensing module directly measures the actual force value and pressure distribution at the clamping interface, realizing precise closed-loop force control. The collaborative work of the vision module and the force sensing module enables the clamp to perceive and respond to dynamic changes during operation in real time, significantly improving operational safety and task success rate.
[0018] 2. The clamping control method of the parallel clamping device based on visual servoing and force feedback described in this invention achieves a seamless transition from coarse positioning to precise contact through the combination of visual guidance and force control switching; online calibration of pre-positioning and visual errors eliminates system positioning errors and improves absolute positioning accuracy; the stable gripping and continuous adaptive process of force-vision fusion enables the clamping process to have closed-loop error correction capability, and can respond to anomalies such as slippage and off-center loading in real time, dynamically adjust the operation strategy, and ensure the reliability and safety of the entire clamping-transfer process. Attached Figure Description
[0019] To make the content of this invention easier to understand, the invention will be further described in detail below with reference to specific embodiments and accompanying drawings.
[0020] Figure 1 This is a schematic diagram of the parallel clamping device based on visual servoing and force feedback in a preferred embodiment of the present invention. Figure 2 This is a schematic diagram of the structure of the three-axis motion platform of the present invention; Figure 3 This is a three-dimensional structural diagram of the clamp drive assembly and the clamp assembly of the present invention; Figure 4 This is an exploded view of the clamp drive assembly and clamp assembly of the present invention. Figure 5 This is a top view of the clamp drive assembly and clamp assembly of the present invention; Figure 6 This is an exploded structural diagram of the clamp assembly and end force sensing module of the present invention; Figure 7 This is a schematic diagram of the three-axis motion platform, vision module, and target object of the present invention; Figure 8 This is a flowchart of the clamping control method of the parallel clamping device of the present invention; Explanation of reference numerals in the accompanying drawings: 1. Three-axis motion platform; 11. X-axis motor; 12. Y-axis motor; 13. Z-axis motor; 2. Clamp drive assembly; 21. Clamp motor; 22. Clamp motor fixed rail; 23. Clamp motor moving rail; 3. Clamp assembly; 31. Fixed clamp; 311. Fixed gripper; 32. Moving clamp; 321. Moving gripper; 33. Fixed clamp mounting block; 34. Moving clamp mounting block; 35. Fixed clamp one; 36. Positioning block; 37. Fixed clamp two; 4. End force sensing module; 5. Vision module; 71. Fastener one; 72. Positioning component one; 73. Fastener two; 74. Positioning component two; 100. Target object. Detailed Implementation
[0021] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, so that those skilled in the art can better understand and implement the present invention. However, the embodiments described are not intended to limit the present invention.
[0022] Reference Figure 1-7 As shown, a parallel clamping device based on visual servoing and force feedback according to the present invention includes: The three-axis motion platform 1 provides translational degrees of freedom for the clamp drive assembly 2. The three-axis motion platform 1 includes an X-axis motor 11, a Y-axis motor 12, and a Z-axis motor 13. The Y-axis motor 12 is mounted on the X-axis motor moving rail, and the Z-axis motor 13 is mounted on the Y-axis motor moving rail, providing translational degrees of freedom along the X, Y, and Z axes for the clamp drive assembly 2. Through the hierarchical arrangement of the X, Y, and Z-axis motors, the clamp is provided with flexible positioning capabilities in three-dimensional space. The structure of the Y-axis motor 13 mounted on the X-axis motor moving rail and the Z-axis motor 12 mounted on the Y-axis motor moving rail forms a stable motion chain, ensuring the precise positioning and smooth movement of the clamp in three-dimensional space, laying the spatial positional foundation for subsequent precision clamping operations.
[0023] The clamp drive assembly 2 provides translational freedom for the clamp assembly 3. The clamp drive assembly 2 is mounted on the moving rail of the Z-axis motor 13 and includes a clamp motor 21, a clamp motor fixed rail 22, a clamp motor moving rail 23, a fixed clamp mounting block 33, and a moving clamp mounting block 34. The fixed clamp mounting block 33 is fixedly connected to the clamp motor fixed rail 22, and the moving clamp mounting block 34 is fixedly connected to the clamp motor moving rail 23. The clamp assembly 3 includes a fixed clamp 31 and a movable clamp 32. The fixed clamp 31 is connected to a fixed clamp mounting block 33, and the movable clamp 32 is connected to a movable clamp mounting block 34. The fixed clamp 31 is provided with a fixed gripper 311, and the movable clamp 32 is provided with a movable gripper 321. The movable gripper 321 is arranged parallel to the fixed gripper 311. An end-effector force sensing module 4 is disposed on the fixed gripper 311 and / or the movable gripper 321 for real-time measurement of clamping force and pressure distribution. The end-effector force sensing module 4 includes a flexible pressure sensing film disposed at the end of the fixed gripper 311 near the movable gripper 321. The flexible pressure sensing film can directly sense the actual pressure distribution at the clamping interface. Compared to traditional indirect measurement methods, this structure eliminates errors in the transmission process, achieving high-precision real-time measurement of clamping force and pressure distribution, and providing a reliable data foundation for precise adjustment in force control mode.
[0024] The vision module 5 is used to acquire the recognition information and pose information of the target object; the vision module 5 is disposed above or to the side of the clamping drive assembly 2, and the vision module 5 includes an industrial camera and a light source, with the optical axis of the industrial camera facing the clamping area of the clamping assembly 3; wherein, the recognition information includes the mass and fragility level of the target object; the pose information includes the three-dimensional pose of the target object.
[0025] The controller, connected to the vision module 5 and the end effector force sensing module 4 respectively, is configured to plan an initial gripping strategy based on the visual information and control the gripping force according to the end effector force sensing information. The controller is configured to switch control modes during gripping based on the sensing information from the end effector force sensing module 4, and to fine-tune the gripping posture and force by fusing real-time visual information and end effector force information to cope with dynamic changes in slippage and off-center loading.
[0026] Specifically, the fixed clamp mounting block 33 is fixedly connected to the clamp motor fixed rail 22 via fastener 71; the movable clamp mounting block 34 is fixedly connected to the clamp motor moving rail 23 via fastener 71; the fixed clamp 31 is fixedly connected to the fixed clamp mounting block 33 via positioning element 74 and fastener 73; and the movable clamp 32 is fixedly connected to the movable clamp mounting block 34 via positioning element 72 and fastener 73.
[0027] like Figure 5 and 6As shown, the fixed clamp 31 includes a first fixed clamp 35, a positioning block 36, and a second fixed clamp 37. The second fixed clamp 37 is equipped with fixed grippers 311. The first fixed clamp 35 and the second fixed clamp 37 are positioned by a second positioning element 74 and then fixedly connected to the fixed clamp mounting block 33. The positioning block 36 is disposed on the first fixed clamp 35 and is used for pre-positioning the target object 100. The separate structure of the first fixed clamp 35 and the second fixed clamp 37 facilitates processing and assembly, and the second positioning element 74 ensures precise alignment of the two. The positioning block 36 enables the pre-positioning function of the target object, maintaining a stable posture before actual clamping, avoiding slippage or deflection during the clamping process, and significantly improving clamping efficiency and success rate.
[0028] An end-effector force sensing module 4 is disposed on the fixed gripper 311 and / or the movable gripper 321 for real-time measurement of clamping force and pressure distribution. The end-effector force sensing module 4 includes a flexible pressure sensing film disposed at the end of the fixed gripper 311 near the movable gripper 321. The flexible pressure sensing film can directly sense the actual pressure distribution at the clamping interface. Compared to traditional indirect measurement methods, this structure eliminates errors in the transmission process, achieving high-precision real-time measurement of clamping force and pressure distribution, and providing a reliable data foundation for precise adjustment in force control mode.
[0029] The vision module 5 is used to acquire the recognition information and pose information of the target object; the vision module 5 is disposed above or to the side of the clamping drive assembly 2, and the vision module 5 includes an industrial camera and a light source, with the optical axis of the industrial camera facing the clamping area of the clamping assembly 3; wherein, the recognition information includes the mass and fragility level of the target object; the pose information includes the three-dimensional pose of the target object.
[0030] The controller 6, connected to the vision module 5 and the end effector force sensing module 4 respectively, is configured to plan an initial gripping strategy based on the visual information and control the gripping force according to the end effector force sensing information. The controller 6 is configured to switch control modes during gripping based on the sensing information from the end effector force sensing module 4, and to fine-tune the gripping posture and force in a coordinated manner by fusing real-time visual information and end effector force information to cope with dynamic changes in slippage and off-center loading.
[0031] In this embodiment, the quality and fragility levels identified in the information provide object attribute parameters for gripping strategy planning, enabling the controller to select an appropriate gripping force range and operating mode based on the object's characteristics. Three-dimensional pose information provides precise geometric data for the clamp's spatial positioning and trajectory planning; the combination of these two aspects achieves personalized and precise gripping operations. The control mode switching function allows the clamp to flexibly select between visual servo or force control modes according to the operation stage, adapting to different task requirements. The fusion and coordinated fine-tuning of visual and force information enables the clamp to simultaneously respond to position deviations and force changes, dynamically adjusting the gripping posture and clamping force, effectively coping with dynamic changes such as slippage and off-center loading, and improving adaptability and operational stability under complex working conditions.
[0032] The controller's collaborative fine-tuning strategy includes a pressure distribution equalization algorithm: when the end force sensing module 4 detects that the force difference between the two sides of the clamp exceeds the threshold ΔF, and the vision module 5 simultaneously detects that the object tilts beyond the threshold Δθ, the controller calculates a small corrective translation δx and drives the clamp to perform this translation while keeping the total force constant, so that the force difference between the two sides returns to a safe range.
[0033] The pressure distribution equalization algorithm uses force-vision fusion to automatically correct the clamp posture while keeping the total clamping force constant, so that the pressure distribution on both sides tends to be balanced. This algorithm avoids the risk of object damage caused by simply increasing the force, realizes adaptive correction of tilted objects, and significantly improves the stability and safety of clamping. It is especially suitable for precision operation of fragile and irregularly shaped objects.
[0034] Example 2 like Figure 8 As shown, the clamping control method based on the parallel clamping device in Embodiment 1 includes the following steps: The visual module 5 acquires the recognition information and pose information of the target object 100. Based on the recognition information and pose information of the target object 100 obtained by the vision module 5, an initial gripping strategy is planned and the gripper is controlled to move to the pre-contact point. After the target is initially fixed by pressing down the positioning block 36 at the fixed clamp 35 end of the visual servo trajectory to the pre-contact point, the force control mode is switched to the force control mode. Based on the real-time feedback of the end force sensing module 4, the moving clamp 32 is slowly closed. When the contact force is detected, the actual contact point is recorded and compared with the visually estimated point to complete the online calibration of visual error. In force control mode, the clamping force is steadily increased to the target value, and the object's pose is monitored simultaneously based on the signal from vision module 5. When uneven pressure distribution and object tilt are detected, visual information and force information are fused together, and a small pose translation or rotation is performed while keeping the total force constant, so as to make the pressure distribution uniform. During the handling process, the clamping force closed-loop stability is maintained and slippage is detected based on the signal from the end force sensing module 4, and the object's pose is monitored based on the signal from the vision module 5. When high-frequency vibration or force change is detected and the object's feature points move, slippage is determined to have occurred, and the clamping force is increased or the posture is adjusted to re-stabilize the object. When an abnormality is detected, the force or pose of the gripper is adjusted by fusing vision and force information.
[0035] This method achieves a seamless transition from coarse positioning to precise contact by combining visual guidance with force control switching; online calibration of pre-positioning and visual errors eliminates system positioning errors and improves absolute positioning accuracy; the stable gripping and continuous adaptive process of force and vision fusion enables the gripping process to have closed-loop error correction capability, and can respond to anomalies such as slippage and off-center loading in real time, dynamically adjust the operation strategy, and ensure the reliability and safety of the entire gripping and handling process.
[0036] Obviously, the above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations here. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.
Claims
1. A parallel clamping device based on visual servoing and force feedback, characterized in that, include: Three-axis motion platform; A clamp drive assembly is mounted on the three-axis motion platform; A clamp assembly is connected to the clamp drive assembly. The clamp assembly includes a fixed clamp and a movable clamp. The fixed clamp is provided with a fixed jaw, and the movable clamp is provided with a movable jaw. The movable jaw is arranged parallel to the fixed jaw. An end force sensing module is installed on the fixed gripper and / or the movable gripper to measure the gripping force and pressure distribution in real time. The vision module is used to acquire the recognition information and pose information of the target object; The controller, which is connected to the vision module and the end force sensing module respectively, is configured to plan an initial gripping strategy based on the vision information and control the gripping force according to the end force sensing information.
2. The parallel clamping device based on visual servoing and force feedback according to claim 1, characterized in that, The three-axis motion platform includes an X-axis motor, a Y-axis motor, and a Z-axis motor. The Y-axis motor is mounted on the X-axis motor moving rail, and the Z-axis motor is mounted on the Y-axis motor moving rail, providing translational degrees of freedom along the X, Y, and Z axes for the clamp drive assembly.
3. The parallel clamping device based on visual servoing and force feedback according to claim 2, characterized in that, The clamp drive assembly is mounted on the Z-axis motor moving rail and includes a clamp motor, a clamp motor fixed rail, a clamp motor moving rail, a fixed clamp mounting block, and a movable clamp mounting block. The fixed clamp mounting block is fixedly connected to the clamp motor fixed rail, and the movable clamp mounting block is fixedly connected to the clamp motor moving rail.
4. The parallel clamping device based on visual servoing and force feedback according to claim 3, characterized in that, The fixed clamp mounting block is fixedly connected to the clamp motor fixed rail by fastener one; the movable clamp mounting block is fixedly connected to the clamp motor movable rail by fastener one; the fixed clamp is fixedly connected to the fixed clamp mounting block by positioning component two and fastener two, and the movable clamp is fixedly connected to the movable clamp mounting block by positioning component one and fastener two.
5. The parallel clamping device based on visual servoing and force feedback according to claim 4, characterized in that, The fixed clamp includes a first fixed clamp, a positioning block, and a second fixed clamp; the first fixed clamp and the second fixed clamp are fixedly connected to the fixed clamp mounting block after being positioned by the second positioning element; the fixed gripper is disposed on the second fixed clamp; the positioning block is disposed on the first fixed clamp and is used to pre-position the target object.
6. The parallel clamping device based on visual servoing and force feedback according to claim 1, characterized in that, The end force sensing module includes a flexible pressure sensing film, which is disposed at the end of the fixed gripper near the moving gripper.
7. The parallel clamping device based on visual servoing and force feedback according to claim 1, characterized in that, The identification information includes the target object's mass and fragility level, and the pose information includes the target object's three-dimensional pose.
8. The parallel clamping device based on visual servoing and force feedback according to claim 1, characterized in that, The controller is configured to switch control modes based on the end force sensing information during the gripping process, and to fine-tune the gripping posture and force by fusing real-time visual information and end force information to cope with dynamic changes in slippage and off-center loading.
9. The parallel clamping device based on visual servoing and force feedback according to claim 8, characterized in that, The controller's collaborative fine-tuning strategy includes a pressure distribution equalization algorithm: when the end force sensing module detects that the force difference between the two sides of the clamp exceeds the threshold ΔF, and the vision module simultaneously detects that the object tilts beyond the threshold Δθ, the controller calculates a small corrective translation δx and drives the clamp to perform this translation while keeping the total force constant, so that the force difference between the two sides returns to a safe range.
10. A gripping control method for a parallel gripper device based on visual servoing and force feedback as described in any one of claims 1-9, characterized in that, Includes the following steps: The vision module acquires the recognition and pose information of the target object. Based on the recognition and pose information of the target object obtained by the vision module, an initial gripping strategy is planned and the gripper is controlled to move to the pre-contact point. After the target is initially fixed by moving to the pre-contact point based on the visual servo trajectory, the force control mode is switched. The moving clamp is slowly closed according to the real-time feedback of the end force sensing module. When the contact force is detected, the actual contact point is recorded and compared with the visually estimated point to complete the online calibration of visual error. In force control mode, the clamping force is steadily increased to the target value, and the object's pose is monitored simultaneously based on visual signals. When uneven pressure distribution and object tilt are detected, visual information and force information are fused together, and a small pose translation or rotation is performed while keeping the total force constant, so as to make the pressure distribution uniform. During the handling process, the clamping force closed-loop stability is maintained and slippage is detected simultaneously based on force sensing signals, and the object pose is monitored based on visual signals; When high-frequency vibration or force change is detected and the object's feature points move, slippage is determined to have occurred, and the clamping force is increased or the attitude is adjusted to re-stabilize the object; when an anomaly is detected, visual and force information are fused to adjust the force or pose of the gripper.