Realization of Primitives by using SVM for Humanoid Robot Walk Generation

Journal Paper

B. Borovac; M. Raković; M. Nikolić

Electronics ISSN: 1450-5843


vol. 15(2) , pp. 19-24


To ensure motion of bipedal humanoid robots in unstructured environment, in general, it is not possible to use preprogrammed trajectories. In this paper is described the approach to generate motion from primitives that ensures on-line modification of locomotor activity. Primitives represent simple movements that are either reflex or learned. Each primitive has its parameters and constraints that are determined on the basis of the movements capable of performing by a human. The set of all primitives represents the base from which primitives are selected and combined for the purpose of performing the corresponding complex movement. SVM regression is used to learn primitives which are verified in the task of walk generation