Based on wavelet package to detect the energy characteristics of fault signals

Based on the results of wavelet package decomposition and the energy variety in different frequency stage,a method can be adopted to detect the faults between the normal signals and the fault ones.If there occur faults in signals,there must be energy variety in different frequency stages.The signal can be decomposed into different frequency stage by wavelet package,and the energy in different stages can be calculated.With the numerical value,an energy vector can be set up,which can express whether the fault occur or not and how severe the faults are.Taken a series of experimental data as an example,the course of the method is explained in detail.The results show that the method can availably distinguish the fault signal between the normal one.This method can be widely used to detect steady and non-steady signals.