人工神经网络在石佛寺滑坡稳定性评价中的应用

综合考虑影响滑坡稳定性的众多因素,建立了基于人工神经网络的滑坡稳定性预测模型,并结合已有的工程实例对所建立的神经网络进行训练,在网络模型的训练过程中,还针对不同训练函数的训练效果进行了对比研究。最后针对湖北省兴山县高阳峡口复建公路段沿线石佛寺Ⅰ、Ⅱ滑坡的稳定性问题进行了预测。结果表明,所建立的滑坡稳定性预测模型具有较好的预测精度,同时也说明了神经网络方法在滑坡稳定性预测中的有效性及良好的发展前景。

APPLICATION OF ARTIFICIAL NEURAL NETWORKS ON STABILITY EVALUATION OF SHIFO-TEMPLE LANDSLIDE

With the consideration of the influencing factors of stability of the slop, the forecasting model based on the artificial neural network is introduced, then the neural network trained by many existent project examples. During the training process, contrasting the results that trainied by different functions. At last, forecast the stability of Shifo-temple Ⅰ、Ⅱ landslides, which locate along the Gaoyang-Xiakou rebuild road that in Xingshan county in Hubei province. Numerical example illustrates that the forecasting model can provide more accurate analyzing results, and also illustrates that the artificial neural network not only have well validity in the forecasting of stability of the slop, but also have nicer development foreground.