Rough-set neural network based fault diagnosis on power-shift-steering-transmission

An oil analysis database of the power-shift-steering-transmission (PSST) on road test was set up. By the rough-set's information reduction and dealing with the imprecise, half-baked, inconsistent data and the neural network's linearity approximate and pattern recognition, we set up the rough-set neural network(RSNN) model. Appling this model on the pattern recognition of the power-shift-steering-transmission and the satisfactory diagnosis results were obtained. This method offers new method and idea on the machinery fault diagnosis for the half-baked information.