由于堆石体介质结构的不均匀性以及堆石体密度与其力学参数之间的非线性关系,因此从实测的力学参数中用线性关系计算密度,很难得到较理想的结果。BP神经网络具有非线性映射、泛化、容错的能力。通过教师信号的学习,能够找到输入与输出之间未知的连续非线性关系,从而达到利用所测力学参数计算密度的目的。 Due to the non-uniformity of the media structure of the rockfill body and the nonlinear relationship between the rockfill density and its mechanical parameters, it is difficult to obtain the ideal result from the linear relationship between the measured mechanical parameters. BP neural network has the ability of nonlinear mapping, generalization and fault tolerance. Through the study of the teacher’s signal, we can find the unknown continuous non-linear relationship between input and output, so as to achieve the purpose of calculating the density using the measured mechanical parameters.