中文摘要：

为研究多种应变速率、不同荷载比例下混凝土强度的变化规律，利用大型混凝土静、动三轴试验系统对混凝土立方体试件进行双向应力状态下的动态试验，在取得的大量试验数据基础上，探讨混凝土动态试验中的数据处理方法．针对混凝土受力状态复杂、其动态强度的影响因素众多等特点，提出利用人工神经网络方法来综合量化多种因素与混凝土动态特性之间的数学关系．通过与传统动态试验数据分析方法进行比较后得出：利用人工神经网络来分析多种因素影响下的混凝土动态特性，是一种简便、适用的新方法，具有良好的应用前景．

英文摘要：

Experimental program on cubic concrete specimens was conducted using a servo - hydraulic multi - axial testing machine to investigate the dynamic properties of concrete in biaxial stress states. The data processing method to deal with the obtained data in the experiment was discussed. Considering the limitations of conventional data processing method, the artificial neural network method was adopted to process the multifactor dynamic properties of concrete. By comparing with the conventional data analysis method, it is proved that the artificial neural network method is more simple and effective in dealing with the complex dynamic properties of concrete and can be used widely in engineering practice.