基于双桥静力触探的长三角全新统冲积层力学参数BP网络预测方法

基于双桥静力触探参数推求区域性土性力学参数的需要,以长三角地区全新统冲积层黏土、粉质黏土、粉土及粉砂为研究对象,通过双桥静力触探参数与螺旋板载荷试验、平板载荷试验、剪切试验及压缩试验的对比分析,认为双桥静探数据可作为土质力学参数预测的评价指标。以此为基础建立长三角全新统冲积层双桥静力触探数据与力学参数的BP神经网络预测模型。最后运用模型对江都泰富港务有限公司原料场双桥静力触探孔各土层力学参数进行计算。结果表明:双桥静探数据可作为长三角全新统冲积层土质力学参数预测的评价指标;建立的力学参数BP神经网络模型具有很高的预测精度,能够满足长三角地区全新统冲积地层的实际工程需求。

Prediction method of cone penetration test of the Yangtze River Delta on holocene alluvial layer mechanical parameters based on BP network

the need of cone penetration parameters for regional soil mechanics parameters based on the Yangtze River Delta, Holocene alluvium clay, silty clay, silt and fine sand as the research object, through the comparison of the cone penetration parameters and spiral plate load test, plate loading test, shear test and compression test analysis, that the evaluation index of static sounding double data can be used as a soil mechanics parameter prediction, and the prediction model of BP neural network based on the Holocene alluvial cone penetration test data and mechanical parameters of the double layer, and finally transported to the Jiangdu Port Company Limited"s raw material field cone penetration hole, the soil mechanical parameters were calculated by the model. The results of the study show that: the bridge static sounding data can be used as evaluation index of the Holocene alluvium soil mechanics parameter prediction; mechanical parameters of the BP neural network model has high prediction accuracy and can meet the practical engineering of Holocene alluvium strata in the Yangtze River Delta region needs.