摘要
对目前广泛应用于加筋土工程的特种筋材——CE131土工网、SDL25土工格栅进行了不同应力水平作用下2年的蠕变试验,获得了大量的实验数据.在此基础上,运用最小二乘法和BP神经网络就其蠕变量进行预测,并将两种预测方法的精度加以对比和分析.
Based on the two-year creep experiments for CE131 geonet and SDL25 geogrid under different load level ,a lot of data are obtained. While the least square and BP neural network are applied to forecast the creep of CE131and SDL25. The least square is a method for finding a curve from the curves to simulate test data. The neural network has the ability of self adaptive learning and remembering. In this paper, the precision of these two methods is compared and analyzed.
出处
《长沙理工大学学报（自然科学版）》
CAS
2007年第1期29-33,共5页
Journal of Changsha University of Science and Technology:Natural Science
基金
湖南省自然科学基金资助项目(01JJY3026)
