【作者】 王晓娟；

【导师】 张泽；

【作者基本信息】 内蒙古大学 ， 模式识别与智能系统， 2008， 硕士

【摘要】 薄板复合材料以其优越的性能而广泛应用于航空航天领域,在使用过程中由于粘接界面脱粘而造成的灾难性事故时有发生,因此亟需对界面粘接质量进行准确的定量检测,以保证产品质量和使用安全。超声脉冲回波法以缺陷定位准确、适用范围广、操作方便等优点广泛应用于无损检测。如何对反映粘接质量的超声检测回波信号进行分类识别,是薄板复合材料超声检测的关键也是难点之一。本文用模糊模式识别方法对薄板粘接质量做出定量识别。首先,利用BP神经网络建立了各个特征量对各粘接质量的模糊隶属函数,实验表明,基于BP神经网络建立的隶属函数即方便简单,又符合实际应用的要求。在此基础上,建立已知粘接质量特征量与未知粘接质量特征量的模糊集,设计模糊模式识别算法,该算法选用海明贴近度和算术平均最小值贴近度作为贴近度函数,然后用择近原则法对薄板粘接质量做出定量识别,实验表明该识别方法是准确有效的。隶属函数的建立是实现模糊模式识别的关键,而模糊模式识别算法应用在薄板复合材料的粘接质量识别中,克服了由复合材料内部结构的复杂性及工作环境变化等因素引起的识别难的问题,为专用数字化超声检测仪的设计和实现奠定了基础。更多还原

【Abstract】 More and more thin composite materials have been applied in industry of aeronautics and astronautics for their superior performance. Disastrous accidents happen in process of working at times for bonding defect of composite material, so it is urgently needed to make accurate detection for bonding interface and insure the quality of products and working safety. Ultrasonic echo method is widely applied in non-destructive detection for its advantage of accurate defect locating, wide application, and convenient operation and so on.How to make classify and recognize echo signal of ultrasonic detection takes information of interface bonding quality is the key and difficulty of ultrasonic detection of thin composite materials. In this paper, we use fuzzy pattern recognition method to judge bonding quality. Firstly, we establish the fuzzy membership function between each characteristic and bonding quality by BP neural network. Simulation results show that this method is convenient, simple, and it conforms to the practical application. Secondly, we establish the fuzzy subsets of each bonding quality and design a fuzzy pattern recognition algorithm. Then we design a fuzzy pattern recognition algorithm to judge bonding quality, and the algorithm selects hamming approach degree and the arithmetic mean least approach degree. Then we use the nearest principle to recognize bonding quality. Simulation results show that the algorithm is very exact for quantitative recognition of bonding quality. The establishing of membership function is the key of fuzzy pattern recognition. Fuzzy pattern recognition algorithm overcomes the problem that caused by the internal structure of composite material and the complexity of the work of environmental and so on. Fuzzy pattern recognition is applied in bonding quality ultrasonic detection of thin composite plate is very helpful for design and realization of digital ultrasonic detection instrument.更多还原