Analog Circuit Fault Diagnosis System Based On Bp Algorithm Dsp

The theory of analog circuit fault diagnosis is very important andsignificative. In this paper, an analog circuit fault diagnosis system isdesigned. Basing on the characteristics of the analog circuit with tolerance,an algorithm using BP neural network is selected for the system, and thisalgorithm is much more mature in theory. This system is implemented on aDSP because the DSP is accuracy and efficient.In the first part of this paper, some algorithms usually used areintroduced, then, the application of BP neural network to the analog circuitfault diagnosis system is discussed, and a program is designed for trainingthe BP neural network by MATLAB. In the second part, a TI cooperation'sDSP(TMS320F2812) is chosen for the diagnosis system based on thecharacteristics of analog circuit;the interface circuit is designed andalgorithm program is tested. At last, an actual circuit is selected to testthe system. The results show that this instrument is much more efficientand accurate than the conventional fault dictionary instrument, and has somepractical value.In this paper, some results are obtained, but there is also somedifference from the practical application,and it still should be studiedin the future.