Algorithmic Design of Low-Power Variable-Stiffness Mechanisms

Abstract

Compliant actuators enabling low-power stiffness adaptation are missing ingredients and key enablers of next generation robotic systems. One of the key components of these actuators is the mechanism implementing stiffness adaptation that requires sophisticated control and nontrivial mechanical design. However, despite recent advances in controlling these systems, their design remains experience based and not well understood. In this paper, we present an optimization-based computational framework for the design of intrinsically low-power compliant variable stiffness mechanisms. The core ingredient of this framework is the mathematical formulation of the design problem—provided by a constrained nonlinear parameter optimization—which is computationally solved here to identify optimal variable stiffness designs. We show the basic capability of this formulation in finding parameters for variable stiffness mechanisms that require the least power by design. Further, we demonstrate the generality of this method in cross-comparing mechanisms with different kinematic topology to identify the one that requires the least power by design.