用人工神经网络模型预测高填方地基工后沉降  

Prediction of the post-load settlement of high embankment by artificial neural network

作　　者：刘宏[1] 李攀峰[2] 张倬元[2] 

机构地区：[1]贵州工业大学喀斯特环境与地质灾害防治重点实验室,贵阳550003 [2]成都理工大学工程地质研究所,成都610059

出　　处：《成都理工大学学报（自然科学版）》2005年第3期284-287,共4页Journal of Chengdu University of Technology: Science & Technology Edition

摘　　要：以人工神经网络模型(多层BP模型)为研究手段,对九寨黄龙机场元山子沟104m的高填方地基工后沉降变形特性进行深入研究。认为高填方地基工后沉降包括填筑过程中产生的部分瞬时沉降、加载结束后的主固结和次固结三部分,瞬时沉降速率最高可达20cm/month,发展时间约为15～40天;主固结速率为0.3～3cm/month,发展时间为8～14个月;次固结速率小于0.3cm/month,发展时间为3～5年。且高填方地基最终沉降量为42～44cm,竣工后半年(2003年4月)工后沉降量完成40%～50%,一年后(2004年4月)可完成70%～80%,两年后(2005年4月)可完成90%～95%。The post-load settlement (occurring after the high embankment attains its scheduled level) of the 104 m high embankment of the Jiuzhaigou-Huanglong airport located in Yuanshanzigou is studied in detail by the artificial neural network (multi-layer BP model). The research results show that the post-load settlement is composedof immediate settlement occurring during the filling process, primary consolidation and secondary consolidation occurring after the loading process. The immediate settlement life time is about 15～40 days with a recorded velocity of 20 cm/month. The primary consolidation velocity is 0.3～3 cm/month, and its life time about 8～14 months. The secondary consolidation velocity is below 0.3 cm/month and its life time about 3～5 years. The maximum post-load settlement is about 42～44 cm and it should take place 40%～50% in half a year (Apr. 2003) after the filling process is over, and reaches 70%～80% in a year (Apr. 2004), and 90%～95% in two years (Apr. 2005).

关 键 词：人工神经网络 高填方地基 工后沉降 预测 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程] TU441.7[自动化与计算机技术—控制科学与工程]