中文摘要：

提出一种基于Daubechies小波和车削面形误差的导轨误差辨识的新方法,根据检测结果的特征分析结果,识别出引起此特征的机床误差源。为了了解小波分析的误差,对工件面形检测信号用不同紧支集的Daubechies小波族在1～9尺度下进行分解和重构,对比重构信号和原始信号发现,紧支集为[0,1]的Db1小波用于工件面形信号分解时误差最小。误差源对加工结果影响的仿真结果与实际加工工件的面形检测结果对比,分离出的表面形貌的单尺度上的信息与原信息除了会有相位时移外,还有很多特征相关性。利用小波的变焦特性,采用特征相关性分析方法,从检测结果中确定出引起加工结果的机床误差源。

英文摘要：

A new method of identifying the guideway errors based on the Daubechies wavelet and workpiece surface of machine tool is proposed.The error causing the character is identified based on the character correction analysis of the test result.In order to demonstrate the analysis error of Daubechies wavelet in signal processing,different compact support Daubechies wavelets were used to resolve the measured flatness error of workpiece.Comparison the reconstructed signals using different compact support Daubechies wavelet shows that decomposition error is the minimum when Daubechies 1-order wavelet of compact support ／[0,1／] is used to process the measured flatness error of workpiece.The separate signal on single scale is with much correlate character except shifted phase.The focal variation character of wavelet and the character correlation analysis are used to identify the machine tool error from the test result.