Fault mode prediction based on decision tree

Chemical industry enterprise is a large scale of production system compounded with complex and large electromechanical devices. As the process industry system, the coupling relationship between each devices are closely related. So the fault diagnosis for this kind of special complex equipment group is very difficult. Equipment failure mode is not easy to predict because of the consistency between the equipment. To adopt the single equipment fault diagnosis methods tend to go wrong, so we must adopt new methods, or change the way we predict. In this paper the model of fault mode prediction is put forward firstly and data classification methods based on decision tree and C4.5 algorithm are studied in this paper. A novel method of fault mode prediction based on decision tree is proposed and the processing is described in detail. The advantage of this prediction method is proved through experiments.