Recognition of state in peg-in-hole by fuzzy schema.

Abstract

The present paper deals with the description and estimation of the task state of a robot by means of fuzzy sets. It is necessary to describe the operation state in order to perform the work by a robot. But in the case where the environment of the operation is uncertain, it is difficult to describe the operation state quantitatively. In this study, we describe this operation state by means of fuzzy schemata which consist linguistically of fuzzy sets and estimate the state by using fuzzy pattern matching. The work object in this study is the peg-in-hole operation of two dimensions. The usefulness of this method is shown by experimental results.