基于神经网络算法土颗粒大小分布的参数模型

Parameter Model for Soil Particle Size Distribution Based on Neural Network Algorithm

公路路基用土都为非饱和土,土水特征曲线定义了非饱和土中吸力与含水率之间的关系。采用Arya-Paris模型可根据土颗粒大小的分布确定非饱和土的土水特征曲线,因此获得土颗粒大小分布的参数模型显得极为重要。提出采用神经网络算法,根据土颗粒分析结果建立土颗粒大小分布的参数模型。工程实例表明,此方法能够满足实际工程精度要求,并具有一定的工程适用性。

The soil used for highway subgrade is unsaturated. The soil-water characteristic curve aennes me relationship between suction and water content of this unsaturated soil, which is confirmed according to the soil particle size distribution with Arya-Paris model. Thus, the obtainment of its parameter model is extremely important. This paper presents a neural network algorithm to establish parameter model of soil particle size distribution by analyzing the result of soil particle test. The engineering examples show that this method can meet the actual accuracy requirements of engineering and has certain engineering applicability.