Fuzzy neurons: a coarse-grained reconfigurable element for computational intelligence

Computational intelligence techniques are increasingly being employed in real-world applications. Accordingly, much design effort is going into developing customizable modules to meet required hardware/software specifications. We have prototyped a fuzzy neuron as coarse-grained reconfigurable element to be used in such a module. This module contains multiple reconfigurable fuzzy neurons, together with built-in memory, interface and fine grained reconfigurable logic to implement a fuzzy neural network in the fashion of a system-on-a-chip. The result is a dynamically reconfigurable computational intelligence based control/decision making system which features a parallel structure and in-situ learning.