Fault Detection and Diagnosis of Analog Circuit Using Soft Computing Technique

Abstract

Circuit fault diagnosis is an important area for electronic manufacturing industries. For critical applications, it is desired to design & develop the system that will detect the fault appearing condition by analyzing various circuit parameters. It is proposed to develop the fault detection and diagnosis system which analyzes circuit parameters like supply current and output voltage of circuit under test (CUT). From these parameters it is possible to identify the deviation in performance of the circuit. The circuit under test is H Bridge Inverter which converts the input DC voltage to AC voltage. Analog circuit fault detection and diagnosis system based on embedded board is achieved. A real time embedded system with ATmega16 microcontroller is developed due to which real time data extraction is possible. Back Propagation neural network algorithm with embedded system is applied to that system. Real-time data collection, analog circuit fault detection and location are designed as the basic functions of fault detection and diagnosis system.