Study on Fault Diagnosis Based on Heterogeneous Integration of Multi-symptom Domains
Hongqiang Wan · Journal of Xi'an Technological University · 2010
In order to make full use of the information of vibration signals of mechanical equipment and avoid the complicated learning,a heterogeneous integration method of multi-symptom domains was put forward.That is,the fault characteristics of different symptom domains were extracted,and heterogeneous classifiers were used to study different symptom domains separately,and their conclusions were integrated.This method makes full use of the redundancy,complementarity and auxiliary of information of different symptom domains and the complementarity of heterogeneous classifiers,so the accuracy and certainty of fault diagnosis were improved.This method was used to diagnose the faults of rotor system.It is shown in the results that heterogeneous integration of multi-symptom domains is better than single classifier and heterogeneous integration of single symptom domain.