HARDWARE IMPLEMENTATION OF ARTIFICIAL NEURAL NETWORKS USING BACK PROPAGATION ALGORITHM ON FPGA

Chaitra.P

Abstract: In order to handle problems such as massive parallelism, Fault tolerance, self learning, adaptivity, computational complexity researchers have developed intelligent system such as artificial neural networks. ANN(Artificial neural network) addresses the issues related to pattern recognition, prediction, associative memory and control. It mimics the human biological neural network and has a human like learning ability and is inspired by its structure, processing method and its learning ability like a human brain. Different algorithms are proposed by the designers to train the neural networks, among those Back propagation algorithm in its gradient descent form is widely used algorithm which provides better performance. Verilog coding is done for ANN and Back propagation training algorithm . The functionality of Verilog is verified by simulation using ModelsimSE 6.3F Simulator. The Verilog code is synthesized using Xilinx ISE 14.7 tool. Finally ANN and Back propagation algorithm was successfully implemented

Keywords: ANN, Back propagation algorithm, Fuzzy logic, Synapses

DOI: https://doi.org/10.15623/ijret.2016.0516044