Quadruped Robot Motion Control via Center-of-Mass Trajectory
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current quadruped robots primarily focus on body structure design and lack effective motion control, limiting them to crawling on four legs, with existing control methods being sensitive to environmental conditions and lacking robustness.
Innovation Solution
A robot motion control method that utilizes a center-of-mass reference trajectory, optimized through an objective function, to generate joint control information, enabling quadruped robots to switch between four-legged and two-wheeled movements, enhancing motion diversity and stability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If quadruped robots focus on body structure design with traditional control methods, then the robot structure is established, but the motion capability is limited to crawling on four legs only
Solution Approach 1:
The patent changes the control parameter from traditional joint-level control to center-of-mass level control. By defining the control objective at the center-of-mass position and attitude, the system can achieve diverse motions (crawling, standing, jumping) through parameter optimization rather than complex structural redesign. The objective function optimizes center-of-mass trajectory to enable different motion modes.
Solution Approach 2:
The control method makes the robot system universal by enabling multiple motion functions (four-legged crawling, two-wheeled standing, jumping) using the same hardware configuration. The center-of-mass control framework provides a unified approach to generate control commands for various motion tasks without requiring additional manipulators or structural modifications.
2Reliability
If traditional control methods are used for quadruped robots, then the control system is simple, but the robot is sensitive to environmental conditions and lacks robustness
Solution Approach 1:
The patent implements feedback control by continuously monitoring the actual center-of-mass position and attitude, comparing them with the reference trajectory, and adjusting control commands accordingly. The objective function incorporates feedback terms that minimize the deviation between actual and desired center-of-mass states, enhancing robustness against environmental disturbances.
Solution Approach 2:
The control method performs preliminary action by pre-planning the center-of-mass reference trajectory before execution. The optimization process anticipates environmental variations and prepares compensatory control commands in advance, allowing the robot to maintain stability and robustness when encountering unexpected environmental conditions during motion execution.
3Productivity
If quadruped robots use conventional crawling motion only, then the control system is simple, but the motion efficiency and flexibility are limited
Solution Approach 1:
The patent applies dynamics by enabling the robot to dynamically switch between different motion modes (crawling, standing, jumping) based on task requirements. The center-of-mass control framework dynamically adjusts the reference trajectory and control commands to optimize motion efficiency for each specific task, allowing high-speed movement when standing on two wheels versus stable crawling on four legs.
Data Source
Figure 1~3(b)
Figure 4~5
Figure 6
AI summary
This application discloses a robot motion control method and apparatus, a robot, and a storage medium, and relates to the field of robot control technologies of artificial intelligence technologies. The method includes: acquiring a center-of-mass reference trajectory used for guiding the robot to execute a target motion; obtaining, by optimizing an objective function, center-of-mass control information for controlling the robot to move while following the center-of-mass reference trajectory; generating joint control information according to the center-of-mass control information and structure information of the robot; and controlling the robot using the joint control information.