A Fault Diagnosis Method Combined Neural Network with Rough Set

The neural network combined with the rough set theory is used to perform fault diagnosis tasks of the self-propelled gun (SPG). We employ a feature extraction algorithm based on rough set to pre-process the raw fault information that would be used by neural network as the training samples. Rough set method can effectively decrease the dimension of the information space. Using this algorithm, the training samples for the neural network can be reduced dramatically, and the training time of the network is decreased. The neural networks adopted were of the feed-forward variety with one hidden layer. They were trained using back-propagation. The method can reduce the false alarm rate and missing alarm rate of the fault diagnosis system effectively, and can detect the composed faults while keep good robustness