Reconfigurable Fuzzy Automaton for Software Agents

This paper presents a reconfigurable architecture to implement the state transitions of the hybrid fuzzy-Boolean finite state machine (HFB FSM) by intelligent software agents. The paper also discusses various aspects of implementing the intelligent software agents in hardware. The reconfigurable hardware accelerator architecture is created using parameterized components to add versatility with respect to the number of fuzzy inputs and the number of states. This is a very attractive property to implement virtual fuzzy automata for supervisory controllers of complex systems in which state clusters vary a great degree. The reconfigurable architecture allows the supervisory controller to reset the fuzzy automaton in order to model a particular state cluster, as needed. The hardware design was done using VHDL.