Gain Scheduling Control for the Hydraulic Actuation of the Hyq Robot Leg

This paper focuses on a theoretical and experimental investigation on the control of a valve actuated hydraulic system driving a biologically inspired robotic leg. The study is part of a more complex project aiming at the design of a power and cognitive autonomous quadruped robot for outdoor operations called HyQ. The first leg prototype has currently two hydraulically activated degrees of freedom (DOF) actuating the hip and knee joints. Hydraulic actuation has been chosen in lieu of electrical, due to its high power-to-weight ratio and fast dynamic response, to meet the specifications in terms of performance and dimensions. The actuation system is composed of a proportional valve and an asymmetric cylinder (one for each DOF). Non-linear and linear system modelling and identification have been undertaken. The measured response in the time domain has been compared with the results of numerical simulation both in the linear and non-linear cases. An initial control implementation has been carried out. A gain scheduling algorithm has been implemented where PID controller gains are adapted according to the joint position within a walking trajectory cycle. The performance of this adaptive controller has been compared with a conventional PID control scheme.