基于人工神经网络方法 ,根据云峰大坝坝顶水平位移观测资料识别大坝混凝土和岩石基础的弹性模量 .采用修正的BP学习算法 ,并通过对迭代步长的优化计算及对观测数据的归一化处理 ,提高了参数识别的速度和精度 .云峰大坝的工程实际应用表明 ,用神经网络方法识别材料参数具有识别精度高和收敛速度快等特性 ,拟合误差小于 0 .15mm . Based on the artificial neural network method, the elastic modulus of dam concrete and rock foundation is identified according to the horizontal displacement observation data of the peak of the dam of Yunfeng Dam. The modified BP learning algorithm is adopted, and the optimal calculation of iteration step length The normalized processing improves the speed and precision of parameter identification.The practical application of Yunfeng Dam shows that the identification of material parameters by neural network has the characteristics of high recognition accuracy and fast convergence, and the fitting error is less than 0.15mm.