ANALOG CIRCUIT-DESIGN AND IMPLEMENTATION OF AN ADAPTIVE RESONANCE THEORY (ART) NEURAL-NETWORK ARCHITECTURE

An analogue circuit implementation is presented for an adaptive resona nce theory neural network architecture, called the augmented ART-1 neu ral network (AARTI-NN). The AARTI-NN is a modification of the popular ARTI-NN, developed by Carpenter and Grossberg, and it exhibits the sam e behaviour as the ART1-NN. The AART1-NN is a real-time model, and has the ability to classify an arbitrary set of binary input patterns int o different clusters. The design of the AARTI-NN circuit is based on a set of coupled nonlinear differential equations that constitute the A ARTI-NN model. The circuit is implemented by utilizing analogue electr onic components such as operational amplifiers, transistors, capacitor s, and resistors. The implemented circuit is verified using the PSpice circuit simulator, running on Sun workstations. Results obtained from the PSpice circuit simulation compare favourably with simulation resu lts produced by solving the differential equations numerically. The pr ototype system developed here can be used as a building block for larg er AARTI-NN architectures, as well as for other types of ART architect ures that involve the AARTI-NN model.