Rule-plus-exception model of knowledge extraction for fault diagnosis of turbine-generator unit
Zhao Xue-zeng · Dianli zidonghua shebei · 2007
An improved rule-plus-exception model of cognitive psychology and machine learning is proposed based on the analysis of fault diagnosis samples regularity and rough set reduction technique,which is suitable for extracting the decision rules from the fault data containing inconsistent information.It is described in detail with its essential structure,and the example of turbine-generator unit fault diagnosis proves its feasibility and availability in which a short list of exceptions are considered,and the sample set is divided into two groups.Comparing the proposed model with the others,its superiority is proved not only on confidence,but also on the generalization ability and succinctness.