摘要
测试了淤泥质软土的流变参数,分析了粘粒含量、含水率等因素对初始剪切力、粘度等流变参数的影响特性和变化规律。根据测试结果和神经网络理论,建立了流变参数BP神经网络模型,确定了模型的结构参数。用该模型预测新样品的流变参数,并与实测数据对比,结果表明预测准确率达90%以上。
The rheological parameters of sullage soft-soil are tested, and the influencing characteristics and changing disciplinarian of clay and water percentage to the initial shear stress and viscosity are analyzed. The BP neural Network model is founded by using the theory of neural network and experiment results, and the parameters of model are given. The rheological parameters of new samples are forecasted by using the model, and the result indicates that the veracity is about 90 % . The results of this research have theoretical signification and practicality value in analyzing the changing disciplinarian of rheological parameters, which is important in rapid forecasting of rheological parameters.
出处
《水文地质工程地质》
CAS
CSCD
北大核心
2007年第2期45-48,共4页
Hydrogeology & Engineering Geology
基金
国家自然科学基金资助项目(50678155)
福建省自然科学基金资助项目(D0410008)
关键词
淤泥质软土
粘粒含量
含水率
流变参数
BP神经网络
sullage soft-soil
clay percentage
water percentage
rheological parameters
BP neural network
