基于径向基神经网络的边坡稳定性评估方法

SIDE SLOPE BASED ON RADIAL BASE FUNCTION NEURAL NETWORK

在现有研究的基础上,引入径向基神经网络理论,提出了边坡稳定性的径向基神经网络预测方法。以土体重度、内摩擦角、粘聚力、锚固段长度等为输入参数,边坡稳定性系数为输出参数,建立精确RBFNN神经网络模型,对边坡稳定性进行了预测,结果表明:用训练成熟的径向基神经网络进行仿真,避免了诸多人为因素的影响,提高了结果的精度,使得计算高效、结果更加准确。

Based on existing researches, radial base function neural network theory is introduced and radial base function neutral network predication method for side slope stability is put forward. Precise RBFNN neural network model is established by taking soil weight, internal friction angle, cohesion and anchorage length as input parameters and slope stability coefficient as output parameter, to predict side slope stability. It is indicated that simulation by using mature radial base function neutral network could avoid many influences caused by human factors and that effective calculation and more accurate result are obtained.