神经网络和模糊综合评判在边坡稳定性分析中的应用比较

Contrastive application of neural network and fuzzy comprehensive evaluation to slope stability analysis

采用BP神经网络和模糊综合评判两种方法应用预测边坡稳定性,选取摩擦角、内聚力、重度、边坡角、孔隙压力比、边坡高度6个指标作为评价因子,结合60个边坡实例,分别采用BP神经网络和模糊综合评判对样本边坡进行稳定性评价.对两者预测结果的优劣性进行评判,结果表明BP神经网络评判结果精度更高,模糊综合评判则可以较好地实现评判等级划分.

The BP neural network and fuzzy comprehensive evaluation were used to predict slope stability and the 6 indexes, such as friction angle, cohesive force, severe, slope angle, pore pressure ratio and slope height were used as the evaluation factors. The stability of the slope was respective evaluated by using BP neural network and fuzzy comprehensive evaluation combining with the examples of 60 slopes. The results show that the accuracy of BP neural network is more accurate, and the fuzzy comprehensive evaluation can be used to achieve the classification of evaluation.