Rough Sets Based Hybrid Intelligent Fault Diagnosis for Precision Test Turntable

Abstract

This paper is concerned with fault diagnosis for the precision test turntable (PTT). Using rough set theory combine with neural network, a forward greedy reduce algorithm based on rough set is presented to pre-process the raw fault information. By calculating the dependence and significance of the condition, the core attributes are gained and finally the reduction of the raw fault information is obtained. The worst case of computational complexity of reduction and the total computational times of the algorithm are presented. The reduced decision table will be used by the neural network as the training samples. Rough set method can effectively decrease the dimension of the information space. In this algorithm, the training samples for the neural network can be reduced dramatically, and the training time of the network is decreased. The method can detect the composed faults while keeping good robustness, and can reduce the false alarm rate and the missing alarm rate of the fault diagnosis system effectively.