Comprehensible Model of a Quasi-periodic Signal

Abstract

In this paper we present a new method to analyze quasi-periodic signals. This method consists of modeling these signals using a Fuzzy Finite State Machine as a particular case of a Linguistic Fuzzy Model of a dynamical system. This model defines states and transitions using a priori knowledge of the signal we want to analyze. The model is represented using fuzzy rules that make it easily comprehensible. We include a practical example analyzing quasi-periodic signals of acceleration measured during the human gait cycle where good results were achieved.