J Intell Robot Syst (2008) 51:3–30
DOI 10.1007/s10846-007-9174-5
Hybrid Dynamic Control Algorithm for Humanoid
Robots Based on Reinforcement Learning
Du´ sko M. Kati ´ c · Aleksandar D. Rodi ´ c ·
Miomir K. Vukobratovi ´ c
Received: 1 September 2006 / Accepted: 31 July 2007 /
Published online: 25 September 2007
© Springer Science + Business Media B.V. 2007
Abstract In this paper, hybrid integrated dynamic control algorithm for humanoid
locomotion mechanism is presented. The proposed structure of controller involves
two feedback loops: model-based dynamic controller including impart-force con-
troller and reinforcement learning feedback controller around zero-moment point.
The proposed new reinforcement learning algorithm is based on modified version of
actor-critic architecture for dynamic reactive compensation. Simulation experiments
were carried out in order to validate the proposed control approach.The obtained
simulation results served as the basis for a critical evaluation of the controller
performance.
Keywords Humanoid robots · Biped locomotion · Integrated dynamic control ·
Reinforcement learning · Actor–critic method
1 Introduction
The contemporary humanoid robots are expected to be servants and maintenance
machines with the main task to assist human activities in our daily life and to replace
humans in hazardous operations. It is as obvious that anthropomorphic biped robots
are potentially capable to effectively move in all unstructured environments where
humans do. There are also strong anticipations that robots for the personal use will
coexist with humans and provide supports such as the assistance for the housework,
D. M. Kati ´ c(B ) · A. D. Rodi ´ c · M. K. Vukobratovi ´ c
Robotics Department, Mihajlo Pupin Institute,
Volgina 15, Belgrade 11060, Serbia
e-mail: dusko@robot.imp.bg.ac.yu
A. D. Rodi ´ c
e-mail: roda@robot.imp.bg.ac.yu
M. K. Vukobratovi ´ c
e-mail: vuk@robot.imp.bg.ac.yu