Constrained optimized command shaping for minimizing residual vibration in a flexible-joint robot
Joint flexibility is a natural trait of robotic manipulators, which limits fast point-to-point motion. Remedial measures are often employed to enable these systems to perform their goal in a desired manner. These measures range from either modifying the system dynamics such that the resonance is increasingly damped or by designing cleverly shaped input commands that avoid exciting the resonant modes altogether. In this work, a numerical framework for generating constrained shaped commands for a two-link flexible-joint robot is presented. To optimally select the design parameters for generating shaped commands, the effects of subjecting the optimization to mutually exhaustive constraints of residual vibration performance, speed of motion and size of actuators has been studied. Few important performance metrics to characterize the performance are also introduced and discussed. The framework has been tested for two basis functions, ramped sinusoid and segmented versine, in simulations and experiments and performance is evaluated against one another and an unshaped bang-bang profile. In practice, it has been shown that the constrained numerical approach reduces vibration in the nonlinear robot system in a more effective and efficient manner than the unconstrained closed-form solution.
Meckl, Purdue University.
Off-Campus Purdue Users:
To access this dissertation, please log in to our