摘要
利用遗传算法和神经网络,基于不同类型,不同条件下非饱和土的吸力测试数据,建立了一种以含水量为主要因素,耦合密度、初始含水量、先期固结压力、孔隙比5个因素的吸力预测模型。预测结果分析表明,所建的模型能很好地拟合试验结果,从而,验证了该模型的合理性和可行性。
By combining genetic algorithm and artificial neural network, an evolutionary neural network model is established for the suction prediction of unsaturated soil. In the model, total density, initial water content, pressure preconsolidated and void ratio are considered. The predicted results, which were obtained through the created model, indicate that this model is well in accord with the test data. It provides a new approach for the research of predicting the shear strength of unsaturated soils.
出处
《岩土力学》
EI
CAS
CSCD
北大核心
2004年第1期73-76,共4页
Rock and Soil Mechanics
基金
教育部跨世纪人才培养计划研究基金资助项目(50179006)。