摘要：The detection and diagnosis of faults are of great practical significance for the safe operation of the plant. The early detection of fault can help avoid system shutdown, breakdown and even catastrophe involving human fatalities and material damage. This paper presents the design and development of artificial neural network based model for the fault detection of differential using a back propagation learning algorithm and multi-layer perceptron neural network. The differential conditions were considered to be healthy differential, bearing fault, worn pinion, worn cranwheel, broken pinion and broken cranwheel which were six neurons of output layer with the aim of fault detection and identification. Features vector which is one of the most significant parameters to design an appropriate neural network were extracted from analysis of acoustic signals in frequency domain by means of Fast Fourier Transform (FFT) method. The statistical features of acoustic signals such as mean, standard deviation, variance, skewness and kourtosis were used as input to ANN. Different neural network structures are analyzed to find the optimal neural network with regards to the number of hidden layers. The results show that the designed system is capable of classifying records with 90.5% accuracy with one hidden layers of neurons in the neural network.
