Hardware implementation of neural network on FPGA for accidents diagnosis of the multi-purpose research reactor of Egypt

摘要

Artificial Neural Networks are applied for solving a wide variety of problems in several areas such as signal processing, robotics, diagnosis, and pattern recognition. These applications demand a high computing power and the traditional software implementation are not sufficient. Hardware implementation of neural networks is very interesting due to its high performance and can easily be made parallel. This paper presents a hardware implementation of neural network after training and simulation on the MATLAB software. The excellent hardware performance has been performed through the use of field programmable gate array (FPGA). The diagnosis of the Multi-Purpose Research Reactor of Egypt accidents is used to test the proposed system.