CN-121973231-A - Collision processing method, device, electronic equipment and computer storage medium

CN121973231ACN 121973231 ACN121973231 ACN 121973231ACN-121973231-A

Abstract

The application provides a collision processing method, a collision processing device, electronic equipment and a computer storage medium. Relates to the technical field of computer science. The method comprises the steps of receiving a control instruction of the double-arm robot, wherein the control instruction is used for indicating the double-arm robot to move to a target space position, determining a collision shell grid corresponding to the double-arm robot and a plurality of target collision pairs according to the control instruction, performing collision judgment on the target space position according to the collision shell grid and the plurality of target collision pairs to generate a collision detection result, and generating a control output signal according to the collision detection result, wherein the control output signal is used for driving the double-arm robot to execute collision-free operation. The method of the application can improve the control reliability of the double-arm robot.

Inventors

YANG XU
Alapati Tursun
LI XILIN
SHI WEN
LI XU
JIN GE

Assignees

北京灵御智能科技有限公司

Dates

Publication Date: 20260505
Application Date: 20260324

Claims (10)

1. A collision processing method, characterized by comprising: Receiving a control instruction of a double-arm robot, wherein the control instruction is used for indicating the double-arm robot to be moved to a target space position; according to the control instruction, determining a collision shell grid and a plurality of target collision pairs corresponding to the double-arm robot; according to the collision shell grid and the plurality of target collision pairs, performing collision judgment on the target space position to generate a collision detection result; and generating a control output signal according to the collision detection result, wherein the control output signal is used for driving the double-arm robot to execute collision-free operation.
2. The method of claim 1, wherein determining a collision housing grid and a plurality of target collision pairs for the two-arm robot in accordance with the control instructions comprises: according to the control instruction, a three-dimensional grid model and a structure description file of the double-arm robot are obtained, wherein the structure description file comprises a robot structure, a connecting rod name and a joint type; Voxel expansion processing is carried out on the three-dimensional grid model, and collision shell grids corresponding to the double-arm robots are generated; and determining a plurality of target collision pairs corresponding to the double-arm robot according to the structure description file.
3. The method of claim 2, wherein voxel expansion processing is performed on the three-dimensional grid model to generate a collision shell grid corresponding to the dual-arm robot, comprising: performing voxel conversion processing on the three-dimensional grid model of the double-arm robot to obtain a three-dimensional voxel matrix of the double-arm robot; Performing expansion treatment on the three-dimensional voxel matrix of the double-arm robot to obtain an expansion Boolean matrix; and performing cube reconstruction on the expansion Boolean matrix to obtain the collision shell grid.
4. The method of claim 2, wherein determining a plurality of target collision pairs corresponding to the dual-arm robot from the structural description file comprises: generating a plurality of initial collision pairs according to the geometric information and joint association relation of each connecting rod of the structure description file; And filtering the plurality of initial collision pairs through three layers of filtering rules to obtain a plurality of target collision pairs, wherein the invalid collision pairs comprise collision pairs with fixed installation contact or geometric constraint exclusion.
5. The method of claim 1, wherein the dual-arm robot comprises a plurality of links, wherein collision determination is performed on the target spatial location based on the collision housing grid and the plurality of target collision pairs, and wherein generating a collision detection result comprises: Determining a pose matrix of each connecting rod corresponding to the geometric body to be detected according to the target space position and the collision shell grid, wherein the pose matrix is used for indicating the space position and the pose of the geometric body to be detected; and performing collision detection processing on the plurality of target collision pairs according to the pose matrix of each connecting rod corresponding to the geometric body to be detected, so as to obtain a collision detection result.
6. The method of claim 5, wherein the plurality of links includes a base link, a left arm link, and a right arm link, and wherein determining a pose matrix for each link for the geometry to be detected based on the target spatial location and the collision shell grid includes: Determining a target radian vector of a left arm connecting rod angle and a target radian vector of a right arm connecting rod angle corresponding to the target space position; updating the initial configuration vector corresponding to the double-arm robot according to the target radian vector of the left arm connecting rod angle and the target radian vector of the right arm connecting rod angle to obtain a target configuration vector; determining a geometric body to be detected corresponding to each connecting rod based on the collision shell grid; And determining the pose matrix of each connecting rod corresponding to the geometric body to be detected according to the target configuration vector.
7. The method of claim 1, wherein generating the control output signal based on the collision detection result comprises: When the collision detection result is collision, the control state of the double-arm robot is adjusted from a normal running state to a collision locking state; in the collision locked state, the control output signal is generated.
8. The method of claim 7, wherein generating the control output signal in the crash-locked state comprises: Judging whether a path from the current position to the target position has collision or not through a joint space interpolation path in the collision locking state; And if the path has no collision, adjusting the control state from a collision locking state to a tracking recovery state, and generating a control output signal as a progressive tracking signal.
9. The method as recited in claim 1, further comprising: And rendering and displaying the global pose, the motion trail of each connecting rod and the collision state of the robot through the visualization tool, and highlighting the detected collision geometric body as a preset color.
10. An electronic device comprising at least one processor and a memory, the memory storing computer-executable instructions, the at least one processor executing the computer-executable instructions stored in the memory, implementing the method of any one of claims 1-9.

Description

Collision processing method, device, electronic equipment and computer storage medium Technical Field The present application relates to the field of computer technologies, and in particular, to a collision processing method, a collision processing device, an electronic device, and a computer storage medium. Background In the scenes of industrial automation, medical surgery, dangerous environment operation (such as nuclear radiation area, deep sea exploration) and remote collaboration, the double-arm robot realizes high-precision and high-flexibility remote task execution through a teleoperation technology. In the prior art, when the double-arm robot is detected to approach other parts of the double-arm robot, the control reduces the movement speed, gives more reaction time to an operator, mainly depends on the operator to avoid the obstacle, is limited by the perception capability and the attention state of the person, and leads to lower control reliability of the double-arm robot. Disclosure of Invention The application provides a collision processing method, a collision processing device, electronic equipment and a computer storage medium, which are used for solving the problem of low control reliability of a double-arm robot in the prior art. In a first aspect, the present application provides a collision processing method, including: Receiving a control instruction of a double-arm robot, wherein the control instruction is used for indicating the double-arm robot to be moved to a target space position; according to the control instruction, determining a collision shell grid and a plurality of target collision pairs corresponding to the double-arm robot; according to the collision shell grid and the plurality of target collision pairs, performing collision judgment on the target space position to generate a collision detection result; and generating a control output signal according to the collision detection result, wherein the control output signal is used for driving the double-arm robot to execute collision-free operation. Optionally, according to the control instruction, determining a collision shell grid and a plurality of target collision pairs corresponding to the dual-arm robot includes: according to the control instruction, a three-dimensional grid model and a structure description file of the double-arm robot are obtained, wherein the structure description file comprises a robot structure, a connecting rod name and a joint type; Voxel expansion processing is carried out on the three-dimensional grid model, and collision shell grids corresponding to the double-arm robots are generated; and determining a plurality of target collision pairs corresponding to the double-arm robot according to the structure description file. Optionally, voxel expansion processing is performed on the three-dimensional grid model, and a collision shell grid corresponding to the double-arm robot is generated, which comprises the following steps: performing voxel conversion processing on the three-dimensional grid model of the double-arm robot to obtain a three-dimensional voxel matrix of the double-arm robot; Performing expansion treatment on the three-dimensional voxel matrix of the double-arm robot to obtain an expansion Boolean matrix; and performing cube reconstruction on the expansion Boolean matrix to obtain the collision shell grid. Optionally, determining, according to the structure description file, a plurality of target collision pairs corresponding to the dual-arm robot includes: generating a plurality of initial collision pairs according to the geometric information and joint association relation of each connecting rod of the structure description file; And filtering the plurality of initial collision pairs through three layers of filtering rules to obtain a plurality of target collision pairs, wherein the invalid collision pairs comprise collision pairs with fixed installation contact or geometric constraint exclusion. Optionally, the dual-arm robot includes a plurality of links, and the collision judgment is performed on the target space position according to the collision shell grid and the plurality of target collision pairs, so as to generate a collision detection result, including: Determining a pose matrix of each connecting rod corresponding to the geometric body to be detected according to the target space position and the collision shell grid, wherein the pose matrix is used for indicating the space position and the pose of the geometric body to be detected; and performing collision detection processing on the plurality of target collision pairs according to the pose matrix of each connecting rod corresponding to the geometric body to be detected, so as to obtain a collision detection result. Optionally, the plurality of links includes a base link, a left arm link, and a right arm link, and determining a pose matrix of each link corresponding to the geometry to be detected according to the target