Optimized Jumping of an Articulated Robotic Leg

Abstract

This paper proposes a nonlinear programming (NLP) formulation intended for the trajectory optimization of legged robot jumping applications during the stance phase, taking into consideration the detailed robot model, actuator capability, terrain condition, etc. The method is applicable to a wide class of jumping robots and was successfully implemented on an articulated robotic leg for jumping in terms of maximum reachable height, minimum energy consumption, as well as optimum energy efficiency. The simulation and experimental results demonstrate that this approach is capable of not only planning one single jumping trajectory, but also designing a periodic jumping gait for legged robots.