An adaptative sampling approach to the problem of learning and control in manipulators systems

  1. Sen Parte, Manuel de la
  2. Sandoval Martín, Juan María
  3. Kryukov, Yury
Journal:
Revista de informática y automática

ISSN: 0210-8712

Year of publication: 1989

Year: 22

Issue: 4

Pages: 31-44

Type: Article

More publications in: Revista de informática y automática

Abstract

An heuristical adaptative sampling approach to the problem of planning, control and learning in manipulator systems is presented. It is based on the use of a hierarchical planning of the control action with the sampling period sequence so as to update this one to the measurable error sequence at sampling points. The control action is performed accordingly to estimation methods (instead to exhaustive calculations) for searching the goal state of the controlled system; local control methods for a multiparametric system with interaction among its elements and local methods for learning for previous movements. The main effort of the paper is devoted to the example of one-link manipulator because it is in the basic structure to obtain models for the multilink case.