Decentralized Model Reference Adaptive Control Design for Modular and Reconfigurable Robots

摘要

Modular and reconfigurable robots (MRRs) offer alluring concepts and have competitive edges over fixed-base manipulators. MRRs consist of joint and link parts assembled into various robot configurations for different types of tasks. The uncertainties in modeling of MRR dynamics and interconnections between joints, in addition to reconfigurability, introduce challenges in control design for MRRs. The objective of this study is developing a hierarchical control design for enhancement of stability and precision in MRR motion control. A low-level decentralized model reference adaptive control (MRAC) design is proposed to be driven by a high-level control structure, which adjusts the robot dynamic model parameters in the equations of motion when the robot configuration is changed. The performance of the proposed control method is verified via simulations.