强夯加固深度预测的神经网络方法研究

摘要

拟用神经网络方法对强夯加固深度进行预测.从国内的若干强夯工程实例中,选取了干重度、孔隙比、主夯击能、夯点间距等强夯加固深度的八个影响因素的相关数据,对相应数据按一定方法分类整理,把这些数据输入到一套基于神经网络方法的FORTRAN程序中对其进行训练,最后用该经过训练程序对两项工程进行加固深度预测.预测结果证明了此种方法不仅有效可行,而且具有工程实用价值和应用前景.%This paper intends to predict the improvement of depth by neural network. From a number of dynamic consolidation project instances of domestic, select the data of the eight factors that influence improvement depth of dynamic consolidation, such as the dry unit weight, void ratio, the main tamping energy, tamping point spacing, and then use the corresponding data which have been processed by certain methods to train a neural network model based on FORTRAN program, at last use the trained program to predict influence improvement of depth of two more projects. The predicted results show that this method not only is feasible and effective, but also has practical value and application prospects.