【摘要】 本文以复合材料常见缺陷分层、孔隙、疏松的超声波检测缺陷信号为研究对象,选取缺陷波的时域及频域特征,考虑到各特征量之间的不可避免地存在耦合关系,而且对缺陷区分的重要性、敏感度也有很大的差别,采用加权多面体隶属度概念构造隶属函数,依据最大隶属原则进行模式识别,从而区分复合材料缺陷类型。更多还原

【Abstract】 Based on signal of compound materials defect such as lamination,hole, loosening in ultrasonic testing,defect signal characteristic parameter in time domain and the frequency domain are selected in the paper.Considering coupling relations between various characteristic parameters and importance and sensitivity in pattern recognition,we structure membership function inclusion weight coefficient according to polyhedron degree of membership concept,and distinguish compound materials defect type by biggest subordination principle.更多还原