Multi-Scale Signal Decomposition And Engineering Surface Quality Research Based On Empirical Mode Decomposition

In machining of parts,surface quality is one of the most specified and important customer requirements.However,the surface quality of the parts such as dimensional accuracy,surface roughness,etc,mainly depends on the cutting tool condition.Tool condition monitoring is an important component of the modern manufacturing technology and has a considerable effect on tool life,product quality,production efficiency and machining operation cost.Starting with indirect method of tool condition monitoring,this paper focuses on multi-scale signals from sensors and three-dimensional engineering surfaces topography,carries out multi-scale signal decomposing and three-dimensional surface filtering,and then establishes the relationship between the tool wear and surface quality.The main research results of the dissertation are as follows:Firstly,starting with the vibration signals,in comparison of Fast Fourier Transform(FT),wavelet transform and empirical mode decomposition(EMD),the paper puts forward an optimal ensemble empirical mode decomposition(EEMD)method for multi-scale signal decomposition.The adaptive algorithm is developed to automatically obtain the initial values of the added white noise amplitude and the ensemble number.The multi-mode search algorithm is explored to optimize the critical two parameters by its performance in global and local search.Thus,the two crucial parameters in EEMD are adaptively obtained.The performances of the proposed method are demonstrated by means of a simulated signal,two bearing vibration signals,and a vibration signal in a milling process.The results show that compared with the traditional EEMD method and other improved EEMD method,the proposed optimal EEMD method automatically obtains the appropriate parameters of EEMD and achieves higher decomposition accuracy and faster computational efficiency.Secondly,taking 3D surface topography of parts using high definition metrology(HDM)as the research object,the bidimensional empirical mode decomposition(BEMD)method is introduced into the 3D surface filtering field,the pater presents a modified fast and adptive bidimensional empirical mode decomposition(MFABEMD)-based separation method.The performances of the proposed method are validated by both a simulated signal and the real-world 3D surface data,and the results that are compared with the current ISO 11562 standardized Gaussian filter demonstrate the proposed method is effective for the separation and extraction of different surface components without the boundary effect.Compared with the original fast and adaptive bidimensional empirical mode decomposition(FABEMD)filter method,the optimal window width of order statistic filter searched by adaptive window algorithm shows its stability and effectiveness,and has higher accuracy and faster time efficiency.Compared with the reprehensive shearlet-based filtering method,the results of the proposed method on surface amplitude parameters are similar.On the other hand,the results on the surface shape and spacing parameters are better than the shearlet-based filtering method.Finally,a modified tool wear measuring method is studied to measure the wear of wiper inserts form experiments.The 3D surface samples are decomposed by the MFABEMD-based filtering method,and 3D surface parameters are calculated of surface roughness.The relationships between the wear mean of the revised axial height of wiper edges,absolute height difference of wiper edges and the average roughness parameter,the root mean square roughness parameter are discussed.Experimental results indicate that the wear mean of the revised axial height of wiper edges shows a strong negative correlation with the average roughness parameter and the root mean square roughness parameter,and the correlation coefficient are-0.9373 and-0.9463,respectively.Meanwhile,absolute height difference of wiper edges shows a strong positive correlation with the average roughness parameter and the root mean square roughness parameter and the correlation coefficient are 0.9334,0.9463,respectively.At last,the relationship between surface quality and tool wear is built,and the paper provides reference for the further research.