Search

CN-122008275-A - Smart hand control method and system

CN122008275ACN 122008275 ACN122008275 ACN 122008275ACN-122008275-A

Abstract

The application provides a smart hand control method and a smart hand control system. The method is applied to electronic equipment, wherein the electronic equipment displays a virtual model corresponding to the smart hand equipment, and the method comprises the steps of detecting a drag operation of a target joint of the virtual model, determining a first joint angle of each joint in the smart hand equipment according to the drag operation, and sending a control instruction to the smart hand equipment, wherein the control instruction comprises the first joint angle of each joint. According to the method, the physical dexterous hand equipment can be directly controlled through the virtual model, so that the control efficiency is improved, and the operation threshold is reduced.

Inventors

  • ZHENG WENJIAN
  • CAI YOUFEI

Assignees

  • 深圳市速腾聚创科技有限公司

Dates

Publication Date
20260512
Application Date
20250923

Claims (12)

  1. 1. A smart hand control method, applied to an electronic device, where the electronic device displays a virtual model corresponding to a smart hand device, the method comprising: Detecting a drag operation for a target joint of the virtual model; Determining a first joint angle of each joint in the smart hand device according to the drag operation; and sending a control instruction to the smart hand equipment, wherein the control instruction comprises a first joint angle of each joint.
  2. 2. The method of claim 1, wherein determining a first joint angle for each joint in the smart hand device from the drag operation comprises: Determining a second joint angle of each joint according to the dragging operation; Performing collision detection according to the second joint angles of the joints to obtain a collision detection result, wherein the collision detection is used for detecting collision contact conditions among the joints of the smart hand equipment or between the smart hand equipment and other objects after the smart hand equipment performs posture adjustment according to the second joint angles of the joints; and determining a first joint angle of each joint according to the second joint angle of each joint based on the collision detection result.
  3. 3. The method of claim 2, wherein the determining the first joint angle of each joint from the second joint angle of each joint based on the collision detection result comprises: In case the collision detection result indicates that there is no collision risk, taking the second joint angle of each joint as the first joint angle of each joint, or And under the condition that the collision detection result indicates that collision risk exists, determining a collision risk index according to normal force or collision time at a collision point, taking the second joint angle of each joint as the first joint angle of each joint under the condition that the collision risk index is smaller than a first threshold value, and carrying out optimization processing on the second joint angle of each joint under the condition that the collision risk index is larger than or equal to the first threshold value to obtain the first joint angle of each joint.
  4. 4. A method according to claim 3, wherein optimizing the second joint angle of each joint to obtain the first joint angle of each joint comprises: determining an attenuation index according to a damping coefficient, collision time and a reverse offset of a collision direction, wherein the damping coefficient is in direct proportion to the environmental rigidity; And performing attenuation compensation on the second joint angle of each joint by using the attenuation index to obtain the first joint angle of each joint.
  5. 5. The method according to claim 2, wherein the performing collision detection according to the second joint angle of each joint, to obtain a collision detection result, includes: The collision detection is carried out by utilizing a directional bounding box detection mode to obtain a first detection result; in the case that the first detection result indicates that there is no bounding box overlapping, taking the first detection result as the collision detection result; Performing collision detection by using a Gilbert-Johnson-Kerrier algorithm to obtain a second detection result when the first detection result indicates that at least two bounding boxes overlap; When the second detection result indicates that the penetration depth of the collision point is smaller than a second threshold value, the second detection result is used as the collision detection result; And carrying out collision detection by using an extended polyhedral algorithm to obtain a collision detection result when the second detection result indicates that the penetration depth of the collision point is greater than or equal to a second threshold.
  6. 6. The method of claim 2, wherein said determining a second joint angle for each joint from said towing operation comprises: Determining a third joint angle of each joint according to a dynamic proportionality coefficient, a displacement vector corresponding to the dragging operation, a rotation axis vector of the target joint and a joint radius of the target joint when the target joint is a non-terminal joint, wherein the dynamic proportionality coefficient is used for performing attenuation compensation on the third joint angle of each joint, the attenuation amplitude is in direct proportion to the proportion of the real-time joint angle of the target joint to the allowed maximum joint angle, and the allowed maximum joint angle of the target joint is in inverse proportion to the environmental rigidity; Determining a second joint angle of each joint based on the third joint angle of each joint and joint constraint conditions of each joint, or Under the condition that the target joint is a tail end joint, the pose of the smart hand device is solved according to a weighted jacobian matrix and a damping coefficient, wherein the weighted jacobian matrix represents the mapping relation between the joint space and the tail end pose space of the smart hand device, in a weight matrix corresponding to the weighted jacobian matrix, the weight values of the thumb, the index finger and the middle finger of the smart hand device are sequentially decreased, and the damping coefficient is in direct proportion to the environmental rigidity; Performing kinematic inverse solution on the pose of the smart hand device to obtain the third joint angle of each joint; And determining the second joint angle of each joint according to the third joint angle of each joint and the joint constraint condition of each joint.
  7. 7. The method according to claim 1, wherein the method further comprises: receiving perception data from the smart hand device, the perception data comprising fourth joint angles of joints on the smart hand device; and updating the gesture of the virtual model according to the perception data.
  8. 8. The method of claim 7, wherein the updating the pose of the virtual model from the perceptual data comprises: Determining a real-time first joint angle deviation according to the temperature and the temperature change rate; Compensating the fourth joint angle of each joint by using the first joint angle deviation to obtain a fifth joint angle of each joint; And updating the posture of the virtual model according to the fifth joint angle of each joint.
  9. 9. The method of claim 7, wherein the updating the pose of the virtual model from the perceptual data comprises: Determining a third joint angle deviation of each joint according to the transmission time delay between the electronic equipment and the smart hand equipment; Compensating the fourth joint angle of each joint by using the third joint angle deviation of each joint to obtain a sixth joint angle of each joint; And updating the posture of the virtual model according to the sixth joint angle of each joint.
  10. 10. The method of claim 9, wherein determining a third joint angular deviation for each joint based on a propagation delay between the electronic device and the smart hand device comprises: Determining a corresponding deviation prediction mode according to the transmission delay; And determining a third joint angular deviation of each joint based on the deviation prediction mode.
  11. 11. The smart hand control system is characterized by comprising electronic equipment and smart hand equipment, wherein the electronic equipment displays a virtual model corresponding to the smart hand equipment; The electronic device is used for detecting the drag operation of a target joint of the virtual model, determining a first joint angle of each joint in the smart hand device according to the drag operation, and sending a control instruction to the smart hand device, wherein the control instruction comprises the first joint angle of each joint; the smart hand equipment is used for receiving the control instruction from the electronic equipment and adjusting the pose of the smart hand equipment according to the first joint angle of each joint in the control instruction.
  12. 12. A computer-readable storage medium storing a computer program, characterized in that the computer program, when executed by a processor, implements the steps of the data adjustment method according to any one of claims 1 to 10.

Description

Smart hand control method and system Technical Field The application relates to the field of robots, in particular to a smart hand control method and a smart hand control system. Background With the development of robotics, various high-precision and high-dexterity mechanical devices are emerging, of which dexterous hand devices are representative. The flexible hand equipment is a high-freedom-degree robot end effector simulating the hand functions of a human, can realize fine grabbing of complex objects through a multi-joint structure, breaks through the limitation of the traditional mechanical clamping jaw, and gives flexibility and adaptability of the robot to approach hands. The dexterous hand equipment has wide application prospect in the fields of minimally invasive surgery, industrial precise assembly and the like. However, how to simply and effectively control smart hand devices is a current problem to be solved. Disclosure of Invention The embodiment of the application provides a smart hand control method and a smart hand control system, which can directly control entity smart hand equipment through a virtual model, improve control efficiency and reduce operation thresholds. In a first aspect, a smart hand control method is provided, and the method is applied to an electronic device, wherein the electronic device displays a virtual model corresponding to the smart hand device, and the method comprises the steps of detecting a drag operation of a target joint of the virtual model; and sending a control instruction to the dexterous hand equipment, wherein the control instruction comprises the first joint angle of each joint. The smart hand control method provided by the invention can convert the dragging operation aiming at the virtual model into the change of the joint angle of the physical smart hand equipment, and the control mode is visual and efficient, reduces the operation threshold and improves the control efficiency. In some embodiments, determining the first joint angle of each joint in the dexterous hand device according to the drag operation comprises determining the second joint angle of each joint according to the drag operation, performing collision detection according to the second joint angle of each joint to obtain a collision detection result, wherein the collision detection is used for detecting collision contact conditions between each joint of the dexterous hand device or between the dexterous hand device and other objects after the dexterous hand device is adjusted according to the second joint angle of each joint, and determining the first joint angle of each joint according to the second joint angle of each joint based on the collision detection result. In some embodiments, determining the first joint angle of each joint from the second joint angle of each joint based on the collision detection result includes taking the second joint angle of each joint as the first joint angle of each joint if the collision detection result indicates that there is no risk of collision. Or determining a collision risk index according to the normal force or the collision time at the collision point under the condition that the collision detection result indicates that the collision risk exists, taking the second joint angle of each joint as the first joint angle of each joint under the condition that the collision risk index is smaller than a first threshold value, and carrying out optimization processing on the second joint angle of each joint under the condition that the collision risk index is larger than or equal to the first threshold value to obtain the first joint angle of each joint. When the user performs a drag operation on the virtual model, the actual motion condition of each joint cannot be perceived, so that the drag operation may cause collision risk of the smart hand device. In the above scheme, the collision risk can be detected, and if the collision risk is too large, the second joint angle of each joint is optimized to reduce the collision risk of the physical joint. In some embodiments, optimizing the second joint angle of each joint to obtain the first joint angle of each joint includes determining an attenuation index according to a damping coefficient, a collision time, and a reverse offset of a collision direction, the damping coefficient being proportional to an environmental stiffness, and performing attenuation compensation on the second joint angle of each joint using the attenuation index to obtain the first joint angle of each joint. The above scheme provides an exponential decay compensation algorithm for trajectory optimization to ensure a smooth transition of the trajectory. And the action of the physical dexterous hand equipment is reversely optimized through the mechanical feedback result (such as collision time, reverse offset of collision direction and the like) of the virtual model, so that the situation that action deviation or equipment damage is easily