A fuzzy finite state machine implementation based on a neural fuzzysystem

摘要

The outputs of a feedforward neural network depend on the presentinputs only. Difficulties arise when a solution requires memory in suchapplications as speech processing, seismic signal processing, languageprocessing, or spatiotemporal signal processing. For such applications,the outputs are not only functions of the present inputs but of thepresent states (or the past inputs and outputs) as well. Fuzzy finitestate machines (FFSMs) can be effectively used in these applications.The aim of this study is to show that a FFSM can be realized using aneural fuzzy system (called NeuFuz). In a FFSM, the output and the nextstate depend on the input and the present state, which in turn is afunction of the previous inputs. To accommodate the memory requirement,the feedforward structure of the neural fuzzy system is changed to arecurrent architecture by adding a feedback loop from the output layerto the input layer during the recall mode. The validity of the approachis verified with a temporal pattern-matching experiment