CN-122008234-A - Robot joint control method and device and robot

CN122008234ACN 122008234 ACN122008234 ACN 122008234ACN-122008234-A

Abstract

The application relates to a robot joint control method and device and a robot. The method comprises the steps of obtaining motion state data of a robot at the current moment, determining joint torque of a target joint in the robot at the next moment according to the motion state data, and controlling movement of the target joint according to the joint torque. The method can accurately control the movement of the robot joint.

Inventors

YUAN JIAWEI
JIANG YIFAN
HUANG XIAO
WANG DONG
WANG XINGFANG
WU QIANG
TAN JIAWEI

Assignees

重庆凤凰技术有限公司

Dates

Publication Date: 20260512
Application Date: 20260327

Claims (13)

1. A method of controlling a robotic joint, the method comprising: acquiring motion state data of the robot at the current moment; Determining joint torque of a target joint in the robot at the next moment according to the motion state data, wherein the target joint comprises a target linear joint and a target rotary joint, the joint torque comprises a linear joint torque when the target joint is the target linear joint, and the joint torque comprises a rotary joint torque when the target joint is the target rotary joint; And controlling the target joint movement according to the joint torque.
2. The method of claim 1, wherein determining joint torque of a target joint in the robot at a next moment in time based on the motion state data comprises: Based on a trained joint torque prediction model, determining joint torque of a target joint in the robot at the next moment according to the motion state data, wherein a hybrid joint mechanical conversion relation corresponding to the target joint is embedded in an output layer of the joint torque prediction model.
3. The method of claim 2, wherein determining joint torque of a target joint in the robot at a next moment in time based on the motion state data comprises: According to the motion state data, determining the expected push rod length of the target push rod corresponding to the target straight joint at the next moment; determining a pushrod thrust of the target pushrod according to the desired pushrod length; And determining the linear joint torque of the target linear joint at the next moment according to the push rod thrust.
4. A method according to claim 3, wherein determining a pushrod thrust for the target pushrod based on the desired pushrod length comprises: Acquiring the current push rod length and push rod expansion speed of the target push rod at the current moment; And determining the push rod thrust of the target push rod according to the current push rod length, the expected push rod length and the push rod telescopic speed.
5. The method of claim 4, wherein determining a pushrod thrust for the target pushrod based on the current pushrod length, the desired pushrod length, and the pushrod telescoping speed comprises: Determining a pushrod travel delta for the target pushrod based on the current pushrod length and the desired pushrod length; And determining the push rod thrust of the target push rod according to the push rod stroke increment and the push rod telescopic speed.
6. A method according to claim 3, wherein the motion state data comprises a joint angle change amount at the present moment of each rotary joint connected to the target straight joint in the robot; According to the push rod thrust, determining the linear joint torque of the target linear joint at the next moment, including: and determining the linear joint torque of the target linear joint at the next moment according to the joint angle change and the push rod thrust.
7. A method according to claim 3, wherein determining a desired pushrod length of a target pushrod corresponding to the target straight joint at the next moment in time based on the motion state data comprises: according to the motion state data, determining pose data of the target straight joint at the next moment; And determining the expected push rod length of the target push rod corresponding to the target straight joint at the next moment according to the pose data.
8. The method of claim 2, wherein determining joint torque of a target joint in the robot at a next moment in time based on the motion state data comprises: Determining an expected joint angle of the target rotary joint at the next moment according to the motion state data; acquiring a current joint angle of the target rotary joint at the current moment; And determining the rotating joint torque of the target rotating joint at the next moment according to the current joint angle and the expected joint angle.
9. The method of claim 8, wherein determining a rotational joint torque of the target rotational joint at the next moment in time based on the current joint angle and the desired joint angle comprises: Determining a joint angle increment of the target rotary joint according to the current joint angle and the expected joint angle; And determining the rotating joint torque of the target rotating joint at the next moment according to the joint angle increment and the joint rotating angular velocity of the target rotating joint at the current moment.
10. The method according to any one of claims 2-9, wherein the joint torque prediction model is iteratively trained by: In each iteration process, acquiring first sample motion state data of a sample robot in a previous iteration process; Determining a sample joint torque of a sample joint in the sample robot in the iterative process according to the first sample motion state data; Controlling the sample joint to move according to the sample joint torque, and generating second sample movement state data of the iterative process; Evaluating the sample joint torque in the iterative process according to the second sample motion state data to obtain a sample evaluation result; And adjusting model parameters of the joint torque prediction model according to the sample evaluation result, wherein sample motion state data of a first iteration process is generated by performing motion control on the sample robot according to preset joint torque, and the sample evaluation result of the first iteration process is obtained by evaluating the preset joint torque.
11. The method of claim 10, wherein the sample motion state data comprises at least one of an actual pitch angle of a sample base in the sample robot, an actual roll angle of the sample base, an actual motion speed of the sample base, an actual motion amplitude of each sample joint, an actual motion rate of each sample joint, a ground contact state of each sample foot of the sample robot: According to the second sample motion state data, evaluating the sample joint torque in the iterative process to obtain a sample evaluation result, including: Determining a base attitude stability evaluation result according to the deviation between the actual pitch angle and the target pitch angle of the sample base and/or the deviation between the actual roll angle and the target roll angle of the sample base; determining a speed tracking evaluation result according to the deviation between the actual movement speed of the sample base and the target movement speed; Determining joint constraint evaluation results according to the deviation between the actual motion amplitude of each sample joint and the target motion amplitude range; Determining an action smoothness evaluation result according to the deviation between the actual motion change rate of each sample joint and the target motion change rate range; determining a foot contact stability evaluation result according to the matching relation between the ground contact state of each sample foot and the target gait phase; and determining the sample evaluation result according to at least one of the base posture stability evaluation result, the speed tracking evaluation result, the joint constraint evaluation result, the action smoothness evaluation result and the foot contact stability evaluation result.
12. A robotic joint control device, the device comprising: the acquisition module is used for acquiring the motion state data of the robot at the current moment; The system comprises a motion state data acquisition module, a determining module, a joint torque acquisition module and a control module, wherein the motion state data acquisition module is used for acquiring motion state data of a robot, the motion state data acquisition module is used for acquiring motion state data of a target joint in the robot, the motion state data acquisition module is used for acquiring motion state data of the robot, and the motion state data acquisition module is used for acquiring motion state data of the robot; and the control module is used for controlling the target joint to move according to the joint torque.
13. A robot comprising a memory and a processor, the memory storing a computer program, characterized in that the processor, when executing the computer program, carries out the steps of the method of any one of claims 1 to 11.

Description

Robot joint control method and device and robot Technical Field The present application relates to the field of robot control technologies, and in particular, to a method and an apparatus for controlling a robot joint, and a robot. Background With the development of technology in the robot field, controlling humanoid robots to stabilize, efficiently and anthropomorphic movements in complex, unstructured dynamic environments is one of the challenges in the robot field. The related art generally controls the motion of a humanoid robot based on a robot dynamics model and a predefined motion trajectory. The above technical solution relies heavily on the accuracy of the robot dynamics model, however, humanoid robots are a highly nonlinear, strongly coupled system with multiple contact dynamics, and it is difficult to obtain a completely accurate model. Therefore, based on the technical scheme, the joint control strategy of the robot cannot be accurately obtained, and the accurate control of the joint motion of the robot cannot be realized. Disclosure of Invention In view of the above, the present application provides a robot joint control method, a robot joint control device, and a robot capable of controlling the movement of a robot joint more accurately. In a first aspect, the present application provides a robot joint control method, including: acquiring motion state data of the robot at the current moment; According to the motion state data, determining joint torque of a target joint in the robot at the next moment, wherein the target joint comprises a target linear joint and a target rotary joint; The target joint motion is controlled based on the joint torque. According to the robot joint control method, based on the motion state data of the robot at the current moment, the linear joint torque of the target linear joint at the next moment and the rotary joint torque of the target rotary joint at the next moment in the robot can be determined, and then the target joint motion of the robot is controlled according to the linear joint torque and the rotary joint torque. In one embodiment, determining the joint torque of the target joint in the robot at the next moment according to the motion state data comprises determining the joint torque of the target joint in the robot at the next moment according to the motion state data based on a trained joint torque prediction model, wherein a hybrid joint mechanical conversion relation corresponding to the target joint is embedded in an output layer of the joint torque prediction model. In this embodiment, according to the joint torque prediction model embedded with the hybrid joint mechanical conversion relationship corresponding to the target joint, the joint torque of the target joint in the robot at the next moment can be more accurately determined, and a foundation is laid for more accurately controlling the joint motion of the robot. In one embodiment, the joint torque of the target joint in the robot at the next moment is determined according to the motion state data, and the method comprises the steps of determining the expected push rod length of the target push rod corresponding to the target straight joint at the next moment according to the motion state data, determining the push rod thrust of the target push rod according to the expected push rod length, and determining the straight joint torque of the target straight joint at the next moment according to the push rod thrust. In this embodiment, according to the motion state data, the expected push rod length of the target push rod corresponding to the target linear joint at the next moment is determined, according to the expected push rod length, the push rod thrust of the target push rod is determined, and according to the push rod thrust, the linear joint torque of the target linear joint at the next moment is determined, that is, the joint torque prediction model outputs the process of the linear joint torque of the target linear joint in the robot at the next moment according to the hybrid joint mechanical conversion relationship corresponding to the target joint, and the process can accurately obtain the linear joint torque of the target linear joint at the next moment without depending on the robot dynamics model, so as to lay a foundation for accurately controlling the movement of the target linear joint in the robot. In one embodiment, determining the push rod thrust of the target push rod according to the expected push rod length comprises obtaining the current push rod length and push rod telescopic speed of the target push rod at the current moment, and determining the push rod thrust of the target push rod according to the current push rod length, the expected push rod length and the push rod telescopic speed. In this embodiment, according to the current push rod length, the expected push rod length and the push rod expansion speed of the target push rod at the current moment, the push rod th