Research on gear-box lubricating fault pattern recognition based on Rough Neural Network

Lubrication fault is difficult to determine in gear-box because of multi-parameters in oil monitoring. A pattern recognition method of gear-box lubricating fault diagnosis based on Rough Neural Network is presented. Considering the multiple faults in complex fault diagnosis system such as wind turbine gear-box, dimension reducing of numerous monitoring parameters by rough set theory is introduced. And a multiple-layer BP neural network is constructed for completing fault pattern recognition. The analytic results reveal that the presented method is effective to composite fault diagnosis in oil analysis. And not only a large quantity of time and energy of the expert can be saved, but efficiency and reliability can be enhanced.