Humanoid Robot Balance Control Under Multi-Task Constraints
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing balance control methods for humanoid robots either waste degrees of freedom or fail to execute multiple tasks due to strong task coupling and inequality constraints, particularly when faced with external disturbances.
Innovation Solution
A balance control method that deconstructs tasks into sole pose, linear momentum, trunk posture, and sole force tasks, while incorporating constraints like dynamic and friction cone constraints, using a multi-task error optimization function to ensure simultaneous task execution.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a simplified model is used for balance control, then the control complexity is reduced, but the degrees of freedom are wasted and multiple tasks cannot be executed simultaneously
Solution Approach 1:
The patent segments the balance control problem into multiple independent task components (e.g., primary tasks and secondary tasks) that can be solved separately and then combined. This allows the robot to execute multiple tasks simultaneously by decomposing the complex control problem into manageable segments, resolving the contradiction between control simplicity and multi-task capability.
Solution Approach 2:
The patent creates a universal balance control framework that can handle multiple different tasks (walking, manipulation, posture maintenance) through a unified optimization approach. This multi-functional control system allows the same control architecture to execute various tasks concurrently, improving versatility without proportionally increasing complexity.
2Adaptability or versatility
If whole body motion control with null-space projection is used to execute multiple tasks, then multiple tasks can be decoupled, but low-priority tasks are sacrificed to ensure high-priority tasks execution
Solution Approach 1:
The patent implements dynamic task prioritization where task priorities are not fixed but can be adjusted in real-time based on system state and task importance. This dynamic approach allows the control system to adaptively allocate degrees of freedom among competing tasks, ensuring that all tasks have a reasonable chance of execution while maintaining overall system reliability.
Solution Approach 2:
The patent incorporates feedback mechanisms that continuously monitor task execution status and system state, allowing the control system to detect when low-priority tasks are failing and adjust the null-space projection accordingly. This feedback loop ensures that task priorities can be dynamically reassigned to prevent task failure, improving overall reliability.
3Adaptability or versatility
If null-space projection method is used to decouple tasks, then high-priority tasks can be executed, but the control system becomes complex and computation time increases
Solution Approach 1:
The patent segments the task hierarchy into distinct levels (primary tasks, secondary tasks, constraints) that can be processed in a structured sequence. This segmentation of the control architecture simplifies the null-space projection process by organizing tasks into manageable groups, reducing computational complexity while maintaining task prioritization capabilities.
Solution Approach 2:
The patent changes the parameter representation of tasks from fixed priority weights to dynamic priority parameters that can be adjusted based on system state. This parameter transformation allows for more efficient computation of null-space projections by using optimized mathematical representations, reducing computational burden while maintaining versatility.
Data Source
AI summary
A humanoid robot balance control method, a humanoid robot, and a storage medium are provided. The method includes: obtaining a task equation of each of a plurality of deconstructed tasks in a corresponding control cycle by solving a plurality of deconstructed task models using a relevant actual state and a corresponding expected state of the humanoid robot; calculating an optimal solution of a multi-task error optimization function based on each task equation; and generating a joint control instruction of the corresponding control cycle based on the optimal solution in response to the optimal solution being obtained within the corresponding control cycle so as to control corresponding joint(s) to execute the tasks. In such manner, it can ensure that the robot satisfies the necessary constraints while executing multiple tasks, and also comprehensively considers the errors of all the tasks to ensure the execution of all the tasks.