软土基坑开挖诱发地表沉降分析  

Intelligent Ground Settlement Analysis Induced by Foundation Pit Excavation in Soft Soil Area

作　　者：孙富学[1] 姜安龙[2] 陈海明[3] 

机构地区：[1]温州大学建筑与土木工程学院,浙江温州325035 [2]南昌航空大学隧道及地下工程研究所,江西南昌330063 [3]岩土及地下工程教育部重点实验室(同济大学),上海200092

出　　处：《南昌航空大学学报（自然科学版）》2009年第1期69-73,共5页Journal of Nanchang Hangkong University(Natural Sciences)

基　　金：温州市科技计划项目"温州市深基坑工程环境效应研究"(S20070041)

摘　　要：基于进化BP神经网络的智能方法,充分利用人工神经网络对非线性函数强大的映射能力、自学习能力和推广能力,文章建立了基于MATLAB神经网络工具箱的优化智能软土基坑开挖诱导地表沉降分析系统,同时采用贝叶斯规则化法对系统进行改善,大幅提高了泛化能力和工程适用性。Ground settlement induced by foundation pit excavation is one of important problems in toundatlon pit engineering in soft soil area. The factors influencing the ground settlement are complex and have stochastic and fuzzy properties, which have brought diffi- culties to the ground settlement analysis. Based on intelligent method of BP artificial neural network, the intelligent ground settlement analysis system is established. For this purpose, powerful functions （Including mapping ability, self- learning ability and generalization ability） of artificial neural network for nonlinear function are utilized, and Bayes rule is used to perfect the system, which results in a remarkable increase in generalization ability and applicability of the system.

关 键 词：基坑开挖 地表沉降 优化智能方法 

分 类 号：TU4[建筑科学—土工工程]