ANALYSES OF THE VETICAL RESISTANCE AND INFLUENCING FACTORS OF THE BOREHOLE CAST-IN-PLACE CONCRETE PILES IN XIAN
Based on the analysis of the mechanical performance of borehole cast-in-place concrete piles, the paper investigates the feasibility of using artificial neural networks to evaluate its vertical resistance. Using the experimental data from the practical projects in Xian, a three-layer back-propagation neural network (BP neural network) is trained to predict the ultimate vertical resistance of the piles. During the network training, 4 important factors are selected as the input while the vertical resistance of the piles is selected as the output. The predicted and measured results show a good agreement on numerical value. Measures to improve the accuracy of the prediction model are discussed. In the further analysis, the influence of evaluation indexes on the vertical resistance of borehole cast- in-place concrete piles is studied.
Artificial neural network borehole cast-in-place concrete piles vertical resistance prediction test data influencing factors
G. Li X. R. Chen
Institute of Architectural Design and Research, Xian University of Architecture and Technology, Xi College of Civil Engineering, Xian University of Architecture and Technology, Xian, China
国际会议
上海
英文
352-359
2007-11-28（万方平台首次上网日期，不代表论文的发表时间）