中文摘要：

混凝土组成复杂，传统的数学方法很难有效预测其各种行为。列出了国内外关于模糊聚类、灰色关联和人工神经网络在混凝土中的应用成果，分析了现代数学方法在定量研究混凝土体系中的不确定关系、确定混凝土性能的主要影响因素、有效预测混凝土的各种性能和优化混凝土配合比等方面的优势，指出今后可在混凝土微观结构方面进一步应用。

英文摘要：

The compositon of concrete was complex, so the traditional mathematical method was difficult to predict its behavior.The main application results of fuzzy clustring, gray relevant and artificial neural networks in concrete at home and abroad were presented.The advantages of modem mathematical methods on the quantitative study of uncertainty relationship in concrete system, determine the main influence factors of concrete performance, and the effective prediction of concrete properties as well as the optimization of the concrete mixture propertions were analyzed.At last,it was pointed out that the modem mathematical methods could be further applied in the studty of concrete microstructure.