Feature Parameters Extraction Of Metal Crack Acoustic Emission Signals Based On DSP

When metal materials appear crack or deformation forced,the strain energy will be released in the form of elastic waves,thus there will be acoustic emission signals.AE signal contains a series of important information of AE sources,that is feature.These features can be simplified to be the form of parameters to identify.AE signals features can be used to study the inner defect of the material and the development trend,to evaluate the material properties and provide meaningful reference for the following operation of equipment and evaluation of safety production.Through analyzing the production mechanism of metal crack,the number of AE signals is changeable in the different stage of metal crack growth.The AE signal has wide range of frequency components. therefore,in order to obtain the feature parameters of metal crack from the AE source,it is very important to preprocess AE signal acquired.According to the needs of subject research,Taking into account the working state of hydraulic turbine,the advantages of digital signal processor(DSP),and the characteristic of acoustic emission signal, Having referred to the related research result,wavelet analysis was chose to de-noise AE signal in advance,then the feature parameters of AE signal were extracted based on DSP.Firstly,this paper summarize the research status of extraction method of AE signal feature parameters,and the advantages and disadvantages of AE signal analysis instrument is analyzed at home and abroad.The conception was put forward that extracting feature parameters of acoustic emission signal based on DSP.The composition and characteristics of DSP Platform was analyzed.the mechanism of wavelet analysis de-Noising and discrete wavelet transform algorithm were studied;then AE signal was processed.According to the definition of the various feature parameters of AE signal,the software design of feature parameters extraction was completed,then,the extraction of feature parameters was realized by using high-speed multi-channels and synchronous AE data collection system,which is controlled and processed by TMS320C6713 chip;Finally,utilizing DSP,the method of extracting feature parameters was verified,respectively using AE signal of breaking off pencil,metal fatigue crack and stretch Fracture.the result is satisfactory.This topic comes from the State Natural Science Fund project 'The Crack Monitor Theory and Method Research of Runner Blade of Francis Turbine.(Project authorization number:50465002).