Diagnostic neural network systems for the electronic circuits

Abstract EN

Neural Networks is one of the most important artificial intelligent approaches for solving the diagnostic processes.

This research concerns with uses the neural networks for diagnosis of the electronic circuits.

Modern electronic systems contain both the analog and digital circuits.

But, diagnosis of the analog circuits suffers from great complexity due to their nonlinearity.

To overcome this problem, the proposed system introduces a diagnostic system that uses the neural network to diagnose both the digital and analog circuits.

So, it can face the new requirements for the modern electronic systems.

A fault dictionary method was implemented in the system.

Experimental results are presented on three electronic systems.

They are: artificial kidney, wireless network and personal computer systems.

The proposed system has improved the performance of the diagnostic systems when applied for these practical cases.