摘要：Differential is one of most important part of power transition system, so condition monitoring and fault diagnosis of this part is very important. In this paper an intelligent method for condition monitoring and fault diagnosis of MF-285 tractor differential, that is one of high product tractors in Iran, has been introduced by acoustic analysis. According to this target first acoustic signals were gathered for health and unhealthy states. After that gathered signals transferred from time domain to frequency domain by fast fourier transfer. After that 29 statistical feature were extracted from signals. Extracted feature were used as entrances to classification of artificial neural network. After mentioned stages accuracy of faults classification system obtained 95.16%, this accuracy expresses that introduced method has high power and quality in condition monitoring and fault diagnosis of differential
