Biped Robot Walking Control on Slopes With Centroid Momentum Planning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current motion control methods for biped robots fail to maintain stable walking on slopes and complex terrains due to their assumption of walking on a plane, leading to poor planning and unstable walking effects, as they cannot correct the angular momentum during each step.
Innovation Solution
A motion control method for biped robots that determines the position of the centroid relative to the swing foot based on state parameters, using preset functions and system models to adjust the angular momentum, allowing for real-time control of the robot's posture and stable walking on uneven terrains.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If conventional motion control methods assume the robot is walking on a plane, then the control system is simpler to implement, but the robot exhibits poor planning ability and unstable walking effect on slopes
Solution Approach 1:
The patent changes the reference frame parameters by introducing a slope coordinate system that rotates with the slope angle, transforming the problem from handling slope disturbances to walking on an equivalent plane. This parameter transformation allows the control system to maintain simplicity while achieving stable walking on slopes by changing how position, velocity, and acceleration are defined and controlled.
Solution Approach 2:
The patent implements dynamic adaptation by continuously estimating the slope angle and adjusting the coordinate system transformation in real-time. The control system dynamically adapts to changing terrain conditions by updating the slope angle estimation and recalculating the transformed coordinates, enabling the robot to maintain stable walking on varying slopes without requiring complex pre-programmed terrain models.
2Ease of operation
If the robot uses fixed trajectory control methods, then the control implementation is straightforward, but the robot cannot correct angular momentum during each step resulting in poor planning ability
Solution Approach 1:
The patent introduces feedback mechanisms by continuously estimating the slope angle from robot state measurements and using this information to dynamically adjust the coordinate transformation and control commands. The feedback loop processes actual robot position, velocity, and acceleration data to correct angular momentum deviations, combining the simplicity of fixed trajectory control with the adaptability of real-time correction.
Solution Approach 2:
The patent applies preliminary action by pre-defining the coordinate transformation relationship based on estimated slope angle, so that when the robot executes control commands, the transformation is already in place. This preliminary setup of the rotating coordinate system allows the robot to naturally correct angular momentum through standard control algorithms without requiring complex real-time calculations during execution.
3Device complexity
If the robot controls centroid position without considering angular momentum adjustment, then the control process is simpler, but the robot cannot achieve stable walking on uneven terrains
Solution Approach 1:
The patent introduces an intermediary coordinate transformation system that acts as a mediator between the simple position control commands and the complex angular momentum requirements. By transforming control commands through the slope-aligned coordinate system, the intermediary layer automatically handles angular momentum compensation without requiring the controller to directly calculate and manage complex rotational dynamics, thus maintaining control simplicity while achieving stability.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A motion control method includes: acquiring (S101) a current state parameter of a robot in response to determining that the robot is in a single-foot supporting state; and determining (S102), based on a first preset function and a preset angular momentum value, the position of the centroid of the robot relative to a swing foot at an end moment of the current single-foot supporting state according to the current state parameter of the robot, such that an angular momentum of the centroid relative to the support foot at an end moment of a next single-foot supporting state reaches the angular momentum value, and controlling (S103) the robot to walk according to the position of the centroid of the robot relative to the swing foot at the end moment of the current single-foot supporting state.