Improving the energy efficiency of bipedal robots with bi-articular actuation

Abstract

The goal of this study is to investigate the effect of actuator position on the theoretical energy consumption of an electrically powered bipedal robot. Specifically, the study focuses on the legs of the afore-mentioned robot because their function is limited, as they are primarily used for either locomotion or standing, and thus, this is suitable for optimization. Furthermore, the hip and knee joints constitute the most powerful joints, and thus, these are studied first. A mathematical formulation is adopted to represent the relationship between the actuators and the robots joints. A genetic optimization is used to minimize the energy loss due to motor winding resistance and no-load torque by altering the position of the actuator with respect to the joints. The results indicate that the energy lost during typical motions can be reduced by up to 30% by using bi-articular actuation. Additionally, most benefits can be realized by adding only a bi-articular actuator between the hip roll and knee pitch axes.

Keywords

• bipedal robots
• bi-articular actuation
• energy efficiency
• design optimization
• generic algorithms