Abstract
In the field of control theory, optimal performance is generally defined as the best possible controlled performance given a static, unchangeable plant system. However, principled design of the underlying system can make designing effective controllers easier and dramatically improve the final control performance beyond what any finely tuned controller could achieve alone. This work develops performance metrics for serial robot arms which help guide the design and optimization of the structure of the arm to achieve greater final performance. First, a kinematic (motion-based) metric called the Actuator Independence Metric (AIM) measures the uniqueness of the movement capabilities of the different joints in a robot arm. Arms which are optimized with respect to the AIM exhibit a greater freedom of movement. In particular, it is shown that the AIM score of a robot correlates strongly with their ability to find solutions to the Inverse Kinematics problem, and that redundant arms with a high AIM score have more useful null-spaces with significant ability to change configuration while maintaining a fixed end-effector pose. Second, a dynamic metric called the Acceleration Radius is explored. The acceleration radius measures the maximum acceleration which a robot arm is capable of generating in any direction. An efficient algorithm for calculating the acceleration radius is developed which exploits the geometry of the mapping from joint torques to acceleration. A design optimization is carried out to demonstrate how the acceleration radius predicts the dynamic movement capabilities of robot arms. It is shown that arms which are optimal with respect to the acceleration radius can follow faster paths through a task space. The metrics developed in this thesis can be used to create customized robot arm designs for specific tasks, which will exhibit desirable control performance.
Degree
MS
College and Department
Ira A. Fulton College of Engineering; Mechanical Engineering
Rights
https://lib.byu.edu/about/copyright/
BYU ScholarsArchive Citation
Morrell, John R., "Designed for Better Control: Using Kinematic and Dynamic Metrics to Optimize Robot Manipulator Design" (2023). Theses and Dissertations. 10058.
https://scholarsarchive.byu.edu/etd/10058
Date Submitted
2023-08-17
Document Type
Thesis
Handle
http://hdl.lib.byu.edu/1877/etd12896
Keywords
serial robot manipulator, design optimization, performance metric, kinematic, dynamic
Language
english