Robust adaptive variable admittance control method based on different intentions in feature space

Abstract

The present application relates to a robust adaptive variable admittance control method based on different intentions in a feature space. The method includes: using the kinematic and dynamic models of the manipulator to determine the mapping relationship between the actual contact force and the virtual contact force; determining the adaptive stiffness coefficient based on the actual contact force change, and determining the robust adaptive virtual damping coefficient based on the direct intention based on the virtual motion speed of the feature point; adjusting the robust adaptive virtual damping coefficient based on the contact force expected by the robot, and determining the admittance control virtual damping coefficient based on direct and indirect intentions; determining the robust adaptive variable admittance control based on human intention in the feature space based on the mapping relationship, the adaptive stiffness coefficient and the admittance control virtual damping coefficient; determining the robot image visual servoing control law based on the real-time image feature points and the expected feature points, obtaining the real-time joint velocity of the manipulator, and controlling the manipulator movement.

Description

Technical Field

[0001] The present application relates to the technical field of robot multi-sensor perception information fusion, and in particular to a robust adaptive variable admittance control method based on different intentions in a feature space. Background Art

[0002] Physical human-machine interaction technologies in industrial and medical scenarios, such as precision assembly, parts inspection, and robotic surgery, often rely on contact information acquired by force sensors. However, most of these systems lack in-depth understanding of specific application scenarios. Furthermore, in complex operating environments, external impact forces often have irreversible effects on robotic control systems, leading to oscillations, instability, and even loss of control, which in turn reduces the system's dynamic performance and operational reliability. Therefore, designing a robust control system with rich sensory information and the ability to effectively cope with external impact disturbances has become key to improving the stability and performance of human-machine collaboration.

[0003] Compliant control is a core technology for enabling safe and natural interactions between robots and their environments or humans, and has been widely applied in robotic contact tasks. Admittance control, a collaborative control method implemented using force sensors and exhibiting a certain degree of robustness, has been widely used in physical human-robot interaction. However, existing control systems generally struggle to balance compliance and robustness. In practical applications, many studies fail to fully consider the impact of external impact forces, resulting in an ineffective balance between compliance and robustness.

[0004] To address this issue, variable admittance control methods based on human intention prediction have been proposed, with some success. However, these methods require complex controller design and are computationally intensive. Furthermore, dampers based on shear-thickening fluids have been used in the field of protective engineering. Their velocity dependence provides excellent compliance and robustness, but these methods typically rely solely on contact force information and lack in-depth environmental awareness. Therefore, robots need to be equipped with visual sensors to achieve more efficient task execution in complex, unstructured environments.

[0005] Visual servo systems, with their exceptional precision, adaptability, and autonomous decision-making capabilities, demonstrate significant advantages in unstructured environments. Incorporating image feedback, visual servo systems can adjust robot posture and path planning in real time and have been widely used in various robotic tasks. However, relying solely on force or visual sensor signals often struggles to meet the demands of complex tasks. Therefore, combining vision with force sensors is an effective approach to addressing diverse tasks. By integrating vision and force information, robots can interact with humans more flexibly, accurately, and safely in dynamic environments. However, because force and visual information are typically at different execution levels in control system design, the system often converges to a local minimum, resulting in inconsistent convergence between force and visual references, which in turn affects system performance.

[0006] Therefore, in related technologies, there is an urgent need for a method that can effectively couple visual and force information in feature space. Summary of the Invention

[0007] Based on this, it is necessary to address the above technical problems and provide a robust adaptive variable admittance control method based on different intentions in the feature space that can effectively couple visual and force information in the feature space.

[0008] In a first aspect, the present application provides a robust adaptive variable admittance control method based on different intentions in a feature space. The method comprises:

[0009] The mapping relationship between actual contact force and virtual contact force is determined using the kinematic and dynamic models of the robot arm;

[0010] determining an adaptive stiffness coefficient based on the actual contact force change, and determining a robust adaptive virtual damping coefficient based on direct intention based on the virtual motion speed of the feature point;

[0011] adjusting the robust adaptive virtual damping coefficient based on the contact force expected by the robot, and determining the virtual damping coefficient for admittance control based on direct and indirect intentions;

[0012] Determining a robust adaptive variable admittance control based on human intention in a feature space based on the mapping relationship, the adaptive stiffness coefficient, and the admittance control virtual damping coefficient;

[0013] Based on the real-time image feature points and expected feature points, the robot image visual servo control law is determined, the real-time joint velocity of the robotic arm is obtained, and the movement of the robotic arm is controlled.

[0014] Optionally, in one embodiment of the present application, determining the mapping relationship between the actual contact force and the virtual contact force using the kinematics and dynamics model of the robotic arm includes:

[0015] Determine the second-order kinematic model of visual servoing based on the robot arm kinematic model and the transformation matrix from the robot arm end effector to the camera;

[0016] The mapping relationship between the actual contact force and the virtual contact force is determined based on the visual servo second-order kinematics model and the robot arm dynamics model.

[0017] Optionally, in one embodiment of the present application, adjusting the robust adaptive virtual damping coefficient based on the contact force expected by the robot and determining the virtual damping coefficient for admittance control based on direct and indirect intentions includes:

[0018] Predict the traction required by robots and humans in real time when collaborating, and determine the desired contact force for adjusting the robots;

[0019] determining a desired virtual admittance parameter based on the contact force desired by the adjustment robot and the robust adaptive virtual damping coefficient;

[0020] A direct and indirect intention based admittance control virtual damping coefficient is determined based on the desired virtual admittance parameter and the robust adaptive virtual damping coefficient.

[0021] Optionally, in one embodiment of the present application, determining and adjusting the desired contact force of the robot includes:

[0022] Compare the actual motion trajectory of the robot end effector with the virtual motion trajectory of the feature point and calculate the trajectory curvature information;

[0023] The desired contact force of the robot is calculated and adjusted based on the curvature information.

[0024] Optionally, in one embodiment of the present application, the virtual damping coefficient of the admittance control based on direct and indirect intentions is:

[0025]

[0026] Among them, D vs is the virtual damping coefficient of admittance control based on direct and indirect intentions, D1 is the robust adaptive virtual damping coefficient based on direct intention, and D expected is the expected virtual admittance parameter.

[0027] Optionally, in one embodiment of the present application, the robust adaptive variable admittance control based on human intention in the feature space is defined as:

[0028]

[0029] Among them, M vs is the virtual mass coefficient in the feature space, D vs is the virtual damping coefficient for admittance control based on direct and indirect intentions, Kvs is the adaptive stiffness coefficient in the characteristic space; e vs is the feature point error; is the acceleration of the feature point error; is the speed of feature point error; s d is the expected feature point position set initially, s ω is the position of the feature point moved by the external contact force, f sext is the virtual contact force in the feature space.

[0030] Optionally, in one embodiment of the present application, the robot image visual servoing control law is defined as:

[0031]

[0032] Among them, v c is the camera's speed, λ is the controller gain, is the pseudo-inverse matrix of the image interaction matrix, e ad is the error between the real-time image feature point and the expected feature point, is the velocity characteristic point, s ω is the position of the feature point moved by the external contact force, s d is the expected feature point position set initially.

[0033] In a second aspect, the present application also provides a robust adaptive variable admittance control device based on different intentions in a feature space. The device includes:

[0034] A mapping relationship determination module is used to determine the mapping relationship between the actual contact force and the virtual contact force using the kinematics and dynamics model of the robot arm;

[0035] a robust adaptive damping coefficient and adaptive stiffness coefficient determination module, configured to determine an adaptive stiffness coefficient based on the actual contact force change, and determine a robust adaptive virtual damping coefficient based on direct intention based on a virtual motion speed of a feature point;

[0036] a direct and indirect intention-based admittance control virtual damping coefficient determination module, configured to adjust the robust adaptive virtual damping coefficient based on the contact force expected by the robot and determine the direct and indirect intention-based admittance control virtual damping coefficient;

[0037] An admittance control determination module, configured to determine a robust adaptive variable admittance control based on human intention in a feature space based on the mapping relationship, the adaptive stiffness coefficient, and the admittance control virtual damping coefficient;

[0038] The robot arm motion control module is used to determine the robot image visual servo control law based on real-time image feature points and expected feature points, obtain the real-time joint speed of the robot arm, and control the movement of the robot arm.

[0039] In a third aspect, the present application further provides a computer device comprising a memory and a processor, wherein the memory stores a computer program and the processor executes the steps of the method described in each of the above embodiments.

[0040] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described in each of the above embodiments.

[0041] The robust adaptive variable admittance control method based on different intentions in the feature space described above first determines the mapping relationship between actual contact force and virtual contact force using the manipulator's kinematic and dynamic models. Next, an adaptive stiffness coefficient is determined based on the actual contact force variation, and a robust adaptive virtual damping coefficient based on direct intention is determined based on the virtual motion velocity of the feature points. The robust adaptive virtual damping coefficient is then adjusted based on the desired contact force, and the virtual damping coefficients for admittance control based on direct and indirect intentions are determined. Finally, robust adaptive variable admittance control based on human intention in the feature space is determined based on the mapping relationship, the adaptive stiffness coefficient, and the virtual damping coefficient for admittance control. Finally, the robot image visual servoing control law is determined based on the real-time image feature points and the desired feature points, and the real-time joint velocity of the manipulator is obtained to control the manipulator's motion. At the direct human intention level, an adaptive virtual damping coefficient is designed in the feature space by analyzing the virtual velocity of the feature points and utilizing the characteristics of the STF. Furthermore, to ensure stable convergence of the manipulator during visual servoing, a virtual stiffness coefficient that is adaptively adjusted based on the contact force variation is designed. These strategies ensure compliance, robustness and stability during human-machine collaborative interaction. At the level of human indirect intention, a curvature-based force guidance method is designed based on the STF virtual damping coefficient. This method compares the actual motion trajectory of the robot's end effector with the virtual motion trajectory of the feature point, and uses the curvature information to calculate the contact force required by the robot. Through this strategy, the robot can accurately follow and guide the human force, thereby achieving smoother and more stable movement, and ensuring robustness and compliance in complex environments. By comparing the actual motion trajectory of the robot's end effector with the virtual motion trajectory of the feature point, the present invention successfully maps the indirect intention curvature model in Cartesian space to the feature space, realizing more accurate force guidance in the process of visual human-machine collaboration, and effectively coordinating the interaction between robots and humans. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 FIG1 is an application environment diagram of a robust adaptive variable admittance control method based on different intentions in a feature space in one embodiment;

[0043] Figure 2 1 is a flow chart of a robust adaptive variable admittance control method based on different intentions in a feature space in one embodiment;

[0044] Figure 3 This is an overall control block diagram in one embodiment;

[0045] Figure 4 Schematic diagram of the composition of a control system in one embodiment;

[0046] Figure 5 A comparison diagram of the motion trajectories of the end effector and the feature points in one embodiment;

[0047] Figure 6 A schematic diagram of a specific experimental verification in an embodiment;

[0048] Figure 7 Schematic diagram of the results of the robustness and convergence verification of the design method when an impact force is applied during the visual servoing process in one embodiment;

[0049] Figure 8 Schematic diagram of experimental results on compliance, robustness, and efficiency of human-machine collaborative motion on a straight trajectory in one embodiment;

[0050] Figure 9 Schematic diagram of experimental results on the compliance, robustness, and efficiency of human-machine collaborative motion on a curved trajectory in one embodiment;

[0051] Figure 10 A schematic diagram illustrating the advantages of the controller designed by the present method in one embodiment compared to a pure visual servoing controller in a hole experiment;

[0052] Figure 11 1 is a structural block diagram of a robust adaptive variable admittance control device based on different intentions in a feature space in one embodiment;

[0053] Figure 12 FIG. 1 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0054] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0055] The robust adaptive variable admittance control method based on different intentions in the feature space provided by the embodiment of the present application can be applied to Figure 1 In the application environment shown, the terminal communicates with the server through the network. The data storage system can store data that the server needs to process. The data storage system can be integrated on the server or placed on the cloud or other network servers. The terminal can be, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can be smart speakers, smart TVs, smart air conditioners, smart car devices, etc. Portable wearable devices can be smart watches, smart bracelets, head-mounted devices, etc. The server can be implemented as a standalone server or a server cluster consisting of multiple servers.

[0056] In one embodiment, Figure 2 As shown in the figure, a robust adaptive variable admittance control method based on different intentions in feature space is provided. Figure 1 The following steps are used as an example to illustrate the server in the example:

[0057] S201: Determine the mapping relationship between the actual contact force and the virtual contact force using the robot arm kinematics and dynamics model.

[0058] In an embodiment of the present application, first, the mapping relationship between the actual contact force of the environment and the virtual contact force in the visual servoing feature space is derived through the kinematics and dynamics model of the robot arm.

[0059] Specifically, in one embodiment of the present application, determining the mapping relationship between the actual contact force and the virtual contact force using the kinematics and dynamics model of the robotic arm includes:

[0060] S301: Determine a second-order kinematic model of visual servoing based on a kinematic model of the robot arm and a transformation matrix from the robot arm end effector to the camera.

[0061] S303: Determine a mapping relationship between actual contact force and virtual contact force based on the visual servoing second-order kinematics model and the robot arm dynamics model.

[0062] In one embodiment of the present application, the relationship between the projection of image feature points in the camera coordinate system onto the 2D image plane of the pinhole camera is:

[0063]

[0064] Among them, γ represents the focal length of the camera, z i Indicates the depth of the image feature point relative to the camera coordinate system.

[0065] The second-order time derivative of the image feature point on the 2D plane is:

[0066]

[0067] Among them, L s is the characteristic Jacobian matrix, which describes the mapping relationship between image feature changes and camera motion. is the derivative of the characteristic Jacobian matrix with respect to time, which represents the dynamic change of the relationship between the feature and the camera motion, V c is the instantaneous speed of the camera, is the time derivative of the camera velocity, that is, the acceleration of the camera.

[0068] The kinematic equation of the robotic arm is:

[0069]

[0070] Among them, x, are the position vector, velocity vector and acceleration vector under the operation room respectively, P(q) is the coordinate system mapping function, q, and are the joint position vector, velocity vector and acceleration vector in the robot joint space, e J e is the Jacobian matrix of the robot arm.

[0071] The second-order kinematic model of visual servoing is determined by the transformation matrix from the end effector of the manipulator to the camera and the kinematic equation of the manipulator:

[0072]

[0073] Among them, L s is the image interaction matrix, c T e is the transformation matrix from the robot end effector to the camera.

[0074] Through the dynamics of the robot arm:

[0075]

[0076] Where M(q) is the inertia matrix of the manipulator, τ is the control input, are the Coriolis matrix and the centripetal force matrix, G(q) is the gravity vector, is the Coulomb force, τ e is the external torque.

[0077] Further calculations yield:

[0078]

[0079] External torque τ e It can be expressed as:

[0080]

[0081] Among them, f e is the force of the end effector of the robot arm, f c is the force in the camera coordinate system, is the transformation matrix from the robot end effector coordinate system to the camera coordinate system.

[0082] Then the virtual contact force in the characteristic space is:

[0083]

[0084] S203: Determine an adaptive stiffness coefficient based on the actual contact force change, and determine a robust adaptive virtual damping coefficient based on direct intention based on the virtual motion speed of the feature point.

[0085] In an embodiment of the present application, in order to ensure different functions in visual servoing and physical human-machine collaboration tasks, a virtual stiffness coefficient is designed to adaptively adjust according to different types of forces (traction or impact force). During the visual servoing process, the virtual stiffness coefficient can ensure the stable completion of the task even under impact force or without external contact. In human-machine collaboration tasks, it can enhance compliance under traction. Experimental analysis of the motion of the UR5 robot shows that the traction force is generally less than 15N, and the incompletely compensated force (set to less than 2N for safety) is managed by the virtual stiffness coefficient. According to this requirement, an adaptive virtual stiffness parameter is designed, which is expressed as follows:

[0086]

[0087] in, f e is the force of the end effector of the robotic arm, s, kg, N, and m are the basic units in the International System of Units, s is the unit of time in seconds, kg is the unit of mass in kilograms, N is the unit of force in Newtons, and m is the unit of length in meters. When the force of the end effector of the robot arm is less than 2N or greater than 15N, the adaptive stiffness coefficient is assigned to an 8×8 diagonal matrix with a value of 200; When the force of the end effector of the robot arm is greater than 2N and less than 15N, the adaptive stiffness coefficient is assigned to an 8×8 diagonal matrix with a value of 0.

[0088] Modeling direct intention plays a crucial role in human-robot collaboration, enabling the robot to capture and infer human behavioral intentions in real time and dynamically adjust control strategies based on this information. By leveraging the unique properties of shear-thickening fluids, an adaptive damping coefficient was designed to account for changes in external impact forces and the traction forces of human-robot collaboration. When the robot arm makes contact with the environment, the virtual force induced by the contact force causes a shift of the desired feature point in feature space. As the contact force increases, the speed of the desired feature point shifts faster. Based on this design, the system dynamically adjusts the damping coefficient based on real-time feedback, precisely controlling the system response and effectively adapting to external disturbances, ensuring stability, compliance, and high-precision control in complex dynamic environments. When there is no contact force on the end-effector, the system remains stable and performs normal servo tasks. When there is contact force on the end-effector, the control system maps the force in Cartesian space to the feature space, forming a virtual mass-damping-stiffness system, which affects the shift of the desired feature point. As the contact force increases, the speed and acceleration of the desired feature point shift faster. The robust adaptive virtual damping coefficient based on direct intention is determined based on the virtual motion velocity of the feature point as follows:

[0089]

[0090] in, μ>0, n>0 are adjustable parameters, is the virtual motion speed of the feature point.

[0091] S205: Adjusting the robust adaptive virtual damping coefficient based on the contact force expected by the robot, and determining the virtual damping coefficient of admittance control based on direct and indirect intentions.

[0092] In an embodiment of the present application, the traction required for collaboration between a robot and a human is predicted in real time, and the admittance parameters are dynamically adjusted, allowing the robot to respond promptly and adjust its control strategy based on the impending action. This estimation, based on indirect intent, enables the robot to more flexibly respond to environmental changes and human action intentions, thereby accurately guiding the robotic arm along a predetermined path. The original robust adaptive virtual damping coefficient based on direct intent is improved by combining indirect intent to determine the virtual damping coefficient for admittance control based on both direct and indirect intent.

[0093] Specifically, in one embodiment of the present application, adjusting the robust adaptive virtual damping coefficient based on the contact force expected by the robot and determining the virtual damping coefficient for admittance control based on direct and indirect intentions includes:

[0094] S401: Predict the traction force required for the robot to collaborate with humans in real time, and determine and adjust the desired contact force of the robot.

[0095] S403: Determine desired virtual admittance parameters based on the contact force desired by the adjustment robot and the robust adaptive virtual damping coefficient.

[0096] S405: Determine a virtual damping coefficient for admittance control based on direct and indirect intentions based on the desired virtual admittance parameter and the robust adaptive virtual damping coefficient.

[0097] In one embodiment of the present application, when a human interacts with a robot and moves it in a specific direction, the goal is to limit the robot's motion in the vertical direction, thereby reducing the interference of error forces. To achieve this, a lower admittance coefficient is selected in the direction of the desired contact force adjusted by the robot, while a higher admittance coefficient is applied in the vertical direction. The higher admittance coefficient results in smaller motion amplitude and slower response in the vertical direction, improving the stability of the system.

[0098] In the feature space, to set higher admittance in the vertical direction and lower admittance in the horizontal direction, the parameters should be changed proportionally with the speed as follows:

[0099]

[0100] Among them, γ1 and γ2 are the admittance parameters in the vertical and horizontal directions respectively. is the estimated velocity of the feature point after being subjected to the force.

[0101] In order to guide the adjustment of the desired contact force of the robot, a higher admittance parameter is set in the direction perpendicular to the force. If the u direction is the direction of the desired contact force of the robot, the desired virtual admittance parameter is set to:

[0102]

[0103] in, The rotation matrix, derived from the desired force component in the u-direction, represents the rotational relationship between the robot base coordinate system and the u-axis coordinate system aligned in the desired direction. By utilizing the rotation matrix R, the admittance controller can adjust parameters along the desired direction of motion, enabling the robot to accurately follow the operator's intent and improve trajectory accuracy and smoothness.

[0104] The rotation matrix R is calculated based on the contact force expected by the adjusted robot, and the specific formula is:

[0105]

[0106] Among them, F v,expected and F u,expected They are used to adjust the components of the robot's desired contact force in the uv direction respectively.

[0107] In the feature space, to ensure that the robot remains stable without external contact force during the visual servoing task, we set the minimum value of the damping coefficient to: In this case, the virtual damping matrix of admittance control is in the eigenspace when

[0108] Then, a virtual damping coefficient for admittance control based on direct and indirect intentions is determined based on the desired virtual admittance parameter and the robust adaptive virtual damping coefficient.

[0109] Specifically, in one embodiment of the present application, the virtual damping coefficient of the admittance control based on direct and indirect intentions is:

[0110]

[0111] Among them, D vs is the virtual damping coefficient of admittance control based on direct and indirect intentions, D1 is the robust adaptive virtual damping coefficient based on direct intention, and D expected is the expected virtual admittance parameter.

[0112] In one embodiment of the present application, determining and adjusting the desired contact force of the robot includes:

[0113] S501: Compare the actual motion trajectory of the robot end effector with the virtual motion trajectory of the feature point, and calculate the trajectory curvature information.

[0114] S503: Calculate and adjust the desired contact force of the robot based on the curvature information.

[0115] In one embodiment of the present application, the curvature of the trajectory of the next cycle can be calculated based on the velocity of the feature point in the current cycle:

[0116]

[0117] in, and is the estimated velocity and acceleration of the feature point in the u direction, and are the estimated velocity and acceleration of the feature point in the v direction.

[0118] The expected contact force of the robot is calculated and adjusted based on the velocity set curvature information of the feature points in the current cycle:

[0119]

[0120] in, represents the centripetal force direction vector, is the tangential driving force direction vector.

[0121] S207: Determine a robust adaptive variable admittance control based on human intention in a feature space based on the mapping relationship, the adaptive stiffness coefficient, and the admittance control virtual damping coefficient.

[0122] In the embodiment of the present application, a robust adaptive variable admittance controller based on human intention in a feature space is obtained through the mapping relationship between different spaces and different human intentions.

[0123] Specifically, in one embodiment of the present application, the robust adaptive variable admittance control based on human intention in the feature space is defined as:

[0124]

[0125] Among them, M vs is the virtual mass coefficient in the feature space, D vs is the virtual damping coefficient for admittance control based on direct and indirect intentions, K vs is the adaptive stiffness coefficient in the characteristic space; e vs is the feature point error; is the acceleration of the feature point error; is the speed of feature point error; s d is the expected feature point position set initially, s ω is the position of the feature point moved by the external contact force, f sext is the virtual contact force in the feature space.

[0126] S209: Determine the robot image visual servo control law based on the real-time image feature points and the expected feature points, obtain the real-time joint speed of the robotic arm, and control the movement of the robotic arm.

[0127] In an embodiment of the present application, the robot image visual servo control law is determined based on the error between the expected feature point and the current feature point of the image of the QR code captured by the camera in real time, and the joint speed of the robotic arm is obtained in real time according to the robot image visual servo control law, thereby controlling the motion of the robotic arm and completing the visual servo process.

[0128] In one embodiment of the present application, the robot image visual servoing control law is defined as:

[0129]

[0130] Among them, v c is the camera's speed, λ is the controller gain, is the pseudo-inverse matrix of the image interaction matrix, e ad is the error between the real-time image feature point and the expected feature point, is the velocity characteristic point, s ω is the position of the feature point moved by the external contact force, sd is the expected feature point position set initially.

[0131] In the above-mentioned robust adaptive variable admittance control method based on different intentions in the feature space, the mapping relationship between actual contact force and virtual contact force is first determined using the manipulator's kinematic and dynamic models. Next, an adaptive stiffness coefficient is determined based on the actual contact force variation, and a robust adaptive virtual damping coefficient based on direct intention is determined based on the virtual motion velocity of the feature points. Next, the robust adaptive virtual damping coefficient is adjusted based on the desired contact force of the robot, and the virtual damping coefficients for admittance control based on direct and indirect intentions are determined. Finally, based on the mapping relationship, the adaptive stiffness coefficient, and the virtual damping coefficient for admittance control, robust adaptive variable admittance control based on human intention in the feature space is determined. Finally, the robot image visual servoing control law is determined based on the real-time image feature points and the desired feature points, and the real-time joint velocity of the manipulator is obtained to control the manipulator's motion. At the direct human intention level, an adaptive virtual damping coefficient is designed in the feature space by analyzing the virtual velocity of the feature points and utilizing the characteristics of the STF. Furthermore, to ensure stable convergence of the manipulator during visual servoing, a virtual stiffness coefficient that is adaptively adjusted based on the contact force variation is designed. These strategies ensure compliance, robustness and stability during human-machine collaborative interaction. At the level of human indirect intention, a curvature-based force guidance method is designed based on the STF virtual damping coefficient. This method compares the actual motion trajectory of the robot's end effector with the virtual motion trajectory of the feature point, and uses the curvature information to calculate the contact force required by the robot. Through this strategy, the robot can accurately follow and guide the human force, thereby achieving smoother and more stable movement, and ensuring robustness and compliance in complex environments. By comparing the actual motion trajectory of the robot's end effector with the virtual motion trajectory of the feature point, the present invention successfully maps the indirect intention curvature model in Cartesian space to the feature space, realizing more accurate force guidance in the process of visual human-machine collaboration, and effectively coordinating the interaction between robots and humans.

[0132] like Figure 3 As shown in the figure, it is the overall control block diagram. Figure 4 Figure 2 shows a schematic diagram of the control system used in a real-world application. The control algorithm and visual processing are implemented on the Ubuntu 18.04LTS operating system. The software framework is based on the VisualServo Platform (ViSP) and the Robot Operating System (ROS).

[0133] like Figure 5The figure below compares the motion trajectories of the end effector and feature points. F.1 to F.4 represent feature points 1 to 4, respectively. It can be observed that the motion trajectory of the robot end effector in the xy direction is consistent with the motion trajectory of the feature points in the uv direction. This demonstrates that the end effector's motion trajectory can be accurately mapped to the motion of the feature points, ensuring that the robot moves along the desired path. It also demonstrates the effective application of the indirect intention method from Cartesian space to feature space, further improving the performance and stability of the admittance control system.

[0134] In one embodiment of the present application, the effect of the present method is verified by a specific experiment, such as Figure 6 As shown, the experiment consists of four parts: 1) Visual servoing experiment: The robot arm moves from its initial position PO to PA. During this process, an impact force is applied to verify the robustness and convergence of the designed servo method; 2) Straight path human-robot collaboration experiment: The end effector is dragged back and forth from position PA to position PB, verifying the compliance and robustness of the human-robot collaboration process; 3) Curved path human-robot collaboration experiment: The end effector is dragged back and forth from position PA to position PB, verifying the compliance and robustness of the human-robot collaboration process; 4) Hole experiment: The coupled admittance control visual servo controller is compared with a pure visual servo controller to demonstrate the superiority of the controller. In the experiment, the shaft diameter is 7 mm and the hole diameter is 8 mm. The workpiece used is a square wooden board with a side length of 20 cm and a hole depth of 2 cm. A 36h11 series AprilTag with a size of 5 cm is affixed to the surface of the board.

[0135] like Figure 7 As shown in the figure, the robustness and convergence of the designed method are verified when the impact force is applied during the visual servoing process. When the impact force is the largest, the convergence speed of this method is faster and the fluctuation of the convergence process is smaller, that is, it has better robustness and convergence.

[0136] like Figure 8 and Figure 9 Figure 2 shows the experimental results for the compliance, robustness, and efficiency of human-robot collaborative motion for linear and curved trajectories. During the experiments, the robot end-effector applied instantaneous impact forces during two human-robot collaborative and two visual servoing processes. The results show that the proposed controller exhibits the best stability, minimal fluctuations during the task, and is able to effectively cope with external impacts, maintain low traction, and complete the task in the shortest time, regardless of linear or curved motion. In contrast, the other controllers exhibited larger fluctuations and poorer stability. Overall, the proposed controller performed well in the compliance and robustness verification, demonstrating increased stability, lower traction, and stronger impact resistance.

[0137] like Figure 10 The figure shows the advantages of the controller designed for this method over a pure visual servo controller in the hole insertion experiment. Experimental analysis demonstrates that, under the designed variable admittance control method, the robot arm can adaptively adjust its virtual mass-spring-damper system regardless of the presence of disturbances. When the end effector initially failed to enter the hole, the system used vision and force feedback to retry the insertion task and ultimately successfully completed it. This demonstrates the system's excellent impact resistance and stability. In contrast, without admittance control, the impact force of the wooden board on the end effector caused the task to fail, and the pure visual servo system entered an emergency shutdown state.

[0138] It should be understood that, although the various steps in the flowcharts involved in the various embodiments described above are displayed in sequence according to the instructions of the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and these steps can be executed in other orders. Moreover, at least a portion of the steps in the flowcharts involved in the various embodiments described above can include multiple steps or multiple stages, and these steps or stages are not necessarily executed and completed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a portion of steps or stages in other steps.

[0139] Based on the same inventive concept, embodiments of the present application further provide a robust adaptive variable admittance control device based on different intentions in a feature space for implementing the robust adaptive variable admittance control method based on different intentions in the feature space mentioned above. The implementation solution provided by this device is similar to the implementation solution described in the above-mentioned method. Therefore, the specific limitations of the embodiments of the robust adaptive variable admittance control device based on different intentions in one or more feature spaces provided below can be found in the limitations of the robust adaptive variable admittance control method based on different intentions in the feature space above, and will not be repeated here.

[0140] In one embodiment, Figure 11 As shown, a robust adaptive variable admittance control device 1100 based on different intentions in a feature space is provided, comprising: a mapping relationship determination module 1101, a robust adaptive damping coefficient and adaptive stiffness coefficient determination module 1103, an admittance control virtual damping coefficient determination module 1105 based on direct and indirect intentions, an admittance control determination module 1107, and a robotic arm motion control module 1109, wherein:

[0141] The mapping relationship determination module 1101 is used to determine the mapping relationship between the actual contact force and the virtual contact force using the kinematics and dynamics model of the robot arm.

[0142] The robust adaptive damping coefficient and adaptive stiffness coefficient determination module 1103 is used to determine the adaptive stiffness coefficient based on the actual contact force change, and determine the robust adaptive virtual damping coefficient based on direct intention based on the virtual motion speed of the feature point.

[0143] The virtual damping coefficient determination module 1105 for admittance control based on direct and indirect intentions is used to adjust the robust adaptive virtual damping coefficient based on the contact force expected by the robot and determine the virtual damping coefficient for admittance control based on direct and indirect intentions.

[0144] The admittance control determination module 1107 is used to determine the robust adaptive variable admittance control based on human intention in the feature space based on the mapping relationship, the adaptive stiffness coefficient and the admittance control virtual damping coefficient.

[0145] The robot arm motion control module 1109 is used to determine the robot image visual servo control law based on the real-time image feature points and the expected feature points, obtain the real-time joint speed of the robot arm, and control the movement of the robot arm.

[0146] In one embodiment of the present application, the mapping relationship determination module is further configured to:

[0147] Determine the second-order kinematic model of visual servoing based on the robot arm kinematic model and the transformation matrix from the robot arm end effector to the camera;

[0148] The mapping relationship between the actual contact force and the virtual contact force is determined based on the visual servo second-order kinematics model and the robot arm dynamics model.

[0149] In one embodiment of the present application, the virtual damping coefficient determination module for admittance control based on direct and indirect intentions is further configured to:

[0150] Predict the traction required by robots and humans in real time when collaborating, and determine the desired contact force for adjusting the robots;

[0151] determining a desired virtual admittance parameter based on the contact force desired by the adjustment robot and the robust adaptive virtual damping coefficient;

[0152] A direct and indirect intention based admittance control virtual damping coefficient is determined based on the desired virtual admittance parameter and the robust adaptive virtual damping coefficient.

[0153] In one embodiment of the present application, determining and adjusting the desired contact force of the robot includes:

[0154] Compare the actual motion trajectory of the robot end effector with the virtual motion trajectory of the feature point and calculate the trajectory curvature information;

[0155] The desired contact force of the robot is calculated and adjusted based on the curvature information.

[0156] In one embodiment of the present application, the virtual damping coefficient of the admittance control based on direct and indirect intentions is:

[0157]

[0158] Among them, D vs is the virtual damping coefficient of admittance control based on direct and indirect intentions, D1 is the robust adaptive virtual damping coefficient based on direct intention, and D expected is the expected virtual admittance parameter.

[0159] In one embodiment of the present application, the robust adaptive variable admittance control based on human intention in the feature space is defined as:

[0160]

[0161] Among them, M vs is the virtual mass coefficient in the feature space, D vs is the virtual damping coefficient for admittance control based on direct and indirect intentions, K vs is the adaptive stiffness coefficient in the characteristic space; e vs is the feature point error; is the acceleration of the feature point error; is the speed of feature point error; s d is the expected feature point position set initially, s ω is the position of the feature point moved by the external contact force, f sext is the virtual contact force in the feature space.

[0162] In one embodiment of the present application, the robot image visual servoing control law is defined as:

[0163]

[0164] Among them, v c is the camera's speed, λ is the controller gain, is the pseudo-inverse matrix of the image interaction matrix, e ad is the error between the real-time image feature point and the expected feature point, is the velocity characteristic point, s ω is the position of the feature point moved by the external contact force, s d is the expected feature point position set initially.

[0165] Each module in the robust adaptive variable admittance control device based on different intentions in the aforementioned feature space can be implemented in whole or in part through software, hardware, or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in the computer device in the form of software, so that the processor can call and execute the corresponding operations of each module.

[0166] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as follows: Figure 11 As shown. The computer device includes a processor, memory, communication interface, display screen and input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner. The wireless manner can be achieved through WIFI, mobile cellular network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, it implements a robust adaptive variable admittance control method based on different intentions in a feature space. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer covering the display screen, or a key, trackball or touchpad provided on the computer device housing, or an external keyboard, touchpad or mouse.

[0167] Those skilled in the art will understand that Figure 11 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0168] In one embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor implements the steps in the above method embodiments when executing the computer program.

[0169] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0170] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0171] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0172] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, database or other media used in the embodiments provided in this application may include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory may include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in the various embodiments provided herein may include at least one of a relational database and a non-relational database. Non-relational databases may include, but are not limited to, distributed databases based on blockchains. The processor involved in the various embodiments provided herein may be, but are not limited to, a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic unit, a data processing logic unit based on quantum computing, and the like.

[0173] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0174] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present application. It should be noted that a person of ordinary skill in the art may make various modifications and improvements without departing from the spirit of the present application, and these modifications and improvements fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be determined by the appended claims.

Claims

1. A robust adaptive variable admittance control method based on different intentions in feature space, characterized by: The method comprises: The mapping relationship between actual contact force and virtual contact force is determined using the kinematic and dynamic models of the robot arm; determining an adaptive stiffness coefficient based on the actual contact force change, and determining a robust adaptive virtual damping coefficient based on direct intention based on the virtual motion speed of the feature point; adjusting the robust adaptive virtual damping coefficient based on the contact force expected by the robot, and determining the virtual damping coefficient for admittance control based on direct and indirect intentions; Determining a robust adaptive variable admittance control based on human intention in a feature space based on the mapping relationship, the adaptive stiffness coefficient, and the admittance control virtual damping coefficient; Determine the robot image visual servo control law based on real-time image feature points and expected feature points, obtain the real-time joint speed of the robotic arm, and control the movement of the robotic arm; The step of adjusting the robust adaptive virtual damping coefficient based on the contact force expected by the robot and determining the virtual damping coefficient for admittance control based on direct and indirect intentions includes: Predict the traction required by robots and humans in real time when collaborating, and determine the desired contact force for adjusting the robots; determining a desired virtual admittance parameter based on the contact force desired by the adjustment robot and the robust adaptive virtual damping coefficient; A direct and indirect intention based admittance control virtual damping coefficient is determined based on the desired virtual admittance parameter and the robust adaptive virtual damping coefficient.

2. The robust adaptive variable admittance control method based on different intentions in feature space according to claim 1 is characterized in that: The mapping relationship between the actual contact force and the virtual contact force is determined by using the kinematics and dynamics model of the manipulator, including: Determine the second-order kinematic model of visual servoing based on the robot arm kinematic model and the transformation matrix from the robot arm end effector to the camera; The mapping relationship between the actual contact force and the virtual contact force is determined based on the visual servo second-order kinematics model and the robot arm dynamics model.

3. The robust adaptive variable admittance control method based on different intentions in feature space according to claim 1 is characterized in that: The determining and adjusting the desired contact force of the robot includes: Compare the actual motion trajectory of the robot end effector with the virtual motion trajectory of the feature point and calculate the trajectory curvature information; The desired contact force of the robot is calculated and adjusted based on the curvature information.

4. The robust adaptive variable admittance control method based on different intentions in feature space according to claim 1, characterized in that: The virtual damping coefficient of the admittance control based on direct and indirect intentions is: in, is the virtual damping coefficient for admittance control based on direct and indirect intentions, is the robust adaptive virtual damping coefficient based on direct intention, is the expected virtual admittance parameter.

5. The robust adaptive variable admittance control method based on different intentions in feature space according to claim 1, characterized in that: The robust adaptive variable admittance control based on human intention in the feature space is defined as: in, is the virtual mass coefficient in the feature space, is the virtual damping coefficient for admittance control based on direct and indirect intentions, is the adaptive stiffness coefficient in the feature space; is the feature point error; is the acceleration of the feature point error; is the speed of feature point error; is the expected feature point position set initially, is the position of the feature point moved by the external contact force, is the virtual contact force in the feature space.

6. The robust adaptive variable admittance control method based on different intentions in feature space according to claim 1, characterized in that: The robot image visual servo control law is defined as: in, is the camera's movement speed, is the controller gain, is the pseudo-inverse matrix of the image interaction matrix, is the error between the real-time image feature point and the expected feature point, is the velocity characteristic point, is the position of the feature point moved by the external contact force, is the expected feature point position set initially.

7. A robust adaptive variable admittance control device based on different intentions in feature space, characterized in that: The device comprises: A mapping relationship determination module is used to determine the mapping relationship between the actual contact force and the virtual contact force using the kinematics and dynamics model of the robot arm; a robust adaptive damping coefficient and adaptive stiffness coefficient determination module, configured to determine an adaptive stiffness coefficient based on the actual contact force change, and determine a robust adaptive virtual damping coefficient based on direct intention based on a virtual motion speed of a feature point; a direct and indirect intention-based admittance control virtual damping coefficient determination module, configured to adjust the robust adaptive virtual damping coefficient based on the contact force expected by the robot and determine the direct and indirect intention-based admittance control virtual damping coefficient; An admittance control determination module, configured to determine a robust adaptive variable admittance control based on human intention in a feature space based on the mapping relationship, the adaptive stiffness coefficient, and the admittance control virtual damping coefficient; The robot arm motion control module is used to determine the robot image visual servo control law based on real-time image feature points and expected feature points, obtain the real-time joint speed of the robot arm, and control the movement of the robot arm; The step of adjusting the robust adaptive virtual damping coefficient based on the contact force expected by the robot and determining the virtual damping coefficient for admittance control based on direct and indirect intentions includes: Predict the traction required by robots and humans in real time when collaborating, and determine the desired contact force for adjusting the robots; determining a desired virtual admittance parameter based on the contact force desired by the adjustment robot and the robust adaptive virtual damping coefficient; A direct and indirect intention based admittance control virtual damping coefficient is determined based on the desired virtual admittance parameter and the robust adaptive virtual damping coefficient.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images