机械加工误差是指零件加工后的实际几何参数（几何尺寸、几何形状和.doc

基于欧几里得贴近度的精密车削误差源模糊诊断方法 王建涛 顾立志* 杜伟文 黄燕华 瞿少魁 华侨大学机电学院 摘 要：机械加工误差直接影响零件的质量和使用性能，机械加工误差来源于加工系统各误差源，误差源的恰当诊断，对控制加工误差具有重要意义。由于误差源的多样性和随机性，其诊断难度很大。提出基于模糊数学原理、采用欧几里得贴近度判别和确定工件加工误差源的方法。加工误差利用模糊子集描述，加工误差源采用语言模糊矢量描述，用七个语言值表达加工误差和加工误差源语言模糊矢量的隶属度；构建加工误差源知识库，得到模糊语言隶属函数的数值表达式，进而得到模糊矩阵；建立加工误差与误差源之间关系的数学模型；计算欧几里得模糊贴近度，结合阈值原则和反向推理，实现加工误差源诊断。以一阶梯轴的加工为例的实验研究表明，本方法对机械加工过程中误差源的诊断准确、快捷。 关键词：欧几里得贴近度；模糊数学；误差源；误差建模；模糊诊断 The Fuzzy Diagnosis Method of Precision Turning Error Sources Based on the Euclidean Closeness Abstract: The Machining errors directly affect the quality and performance of parts, machining errors are derived from the various error sources of processing system. Proper diagnosing of the error sources has great significances to control machining errors. Due to the diversity and randomicity of error sources, the diagnosis has a great difficulty. Based on fuzzy mathematics principle, the method of the Euclidean closeness is put forward to discriminate and determine the machining error sources. The machining errors is described with fuzzy subset and machining error sources are described by language fuzzy vector, and the membership is expressed with seven language value. The knowledge base of machining error sources is constructed, attaining the numeric expression of fuzzy language membership functions, and then fuzzy matrix is obtained. Mathematical model of the relation between machining errors and error sources is established. Then Euclidean fuzzy closeness is calculated to realize the machining error sources diagnosis, combining threshold principle and backward reasoning. An example study of a stepped shaft shows that this method is accurate and fast to machining error sources diagnosis. Keywords: Euclidean closeness; Fuzzy mathematics; Error sources; Error modeling; Fuzzy diagnosis 机械加工误差是指加工后获得的零件的实际几何参数（包括几何尺寸、几何形状及表面间相互位置关系）和理想几何参数之间偏离的程度。机械加工过程是一个不断变化的极其复杂的过程，对零件加工精度有影响的因素很多，且各种影响因素与产品精度之间对应关系复杂，往往是非确定的、模糊的。机械加工误差直接影响零件的质量和使用性能，机械加工误差来源于加工系统各误差源。对误差源的恰