Identification of discrete event systems using the compound recurrent neural network: Extracting DEVS from trained network

Abstract

The authors consider identifying an unknown discrete event system (DES) as recognition of characteristic functions of a discrete event systems specification (DEVS) model that validly represents the system. Such identification consists of two major steps: behavior learning using a specially designed neural network and extraction of a DEVS model from the learned neural network. This paper presents a method for extracting a DEVS model from one such neural network called CRNN (compound recurrent neural network), which is trained using observed input/output events of an unknown DES. The DES to be identified is restricted to a subclass of DES in which any unknown state can be determined by a finite number of input/output sequences. Identification experiments were performed with three types of unknown DESs, the result of which verified the validity of the proposed model extraction method.