摘要
为获得一种实时判别复合材料结构损伤的方法研究了属性测度及其识别。用材料结构的应力变化和表面短裂纹群作为特征矢量对结构的状态进行离散化处理 ,将属性测度识别方法应用于各离散状态的识别 ,从而对材料结构的损伤程度进行评估。实验表明这种方法的综合识别效率优于径向基函数神经网络 ,为研究复合材料结构的损伤在线识别提供了一种研究方法。
This paper considered the attribute measurement and attribute measurement recognition to obtain a real-time damage-detecting method for composite materials and structure. The structure interior strain and structure surface short cracks both acts as state eigenvector to discrete the continuous state, then the attribute measurement recognition is applied to distinct them, therefore composite structure damage level assessment is conducted. The experiment result upon specimen indicates that the synthetic performance of the method is superior to RBF ANN. So an additional approximation of structure damage detection online was proposed.
出处
《材料科学与工程学报》
CAS
CSCD
北大核心
2004年第5期666-668,共3页
Journal of Materials Science and Engineering
基金
国家自然科学基金资助项目 (50 1 350 30
90 2 0 50 31 )
关键词
属性集
属性测度
复合材料结构
attribute set
attribute measurement
composite material and structure
