| 基于遗传-神经网络的无粘结部分预应力高强混凝土梁延性预测模型 |
| |
| 引用本文： | 李哲,胡立黎. 基于遗传-神经网络的无粘结部分预应力高强混凝土梁延性预测模型[J]. 西安理工大学学报, 2005, 21(3): 281-284 |
| |
| 作者姓名： | 李哲 胡立黎 |
| |
| 作者单位： | 西安理工大学,水利水电学院,陕西,西安,710048 |
| |
| 摘 要： | 尝试利用遗传-神经网络模型,对无粘结部分预应力高强混凝土梁的延性进行预报.选用23个试件为学习样本,3个试件为测试样本.模型得到了很好的预测结果.模型中使用了最优保存策略和实数编码.同时,为了防止＂早熟＂现象的发生,提出了一种新的改变适应度值的方法.
|
| 关 键 词： | 遗传算法 神经网络 高强混凝土梁 预应力 延性 |
| 文章编号： | 1006-4710(2005)03-0281-04 |
| 收稿时间： | 2005-03-21 |
| 修稿时间： | 2005-03-21 |
Prediction Model of Ductility of Unbonded Partially Prestressed Concrete Beam with High Strength Based on Genetic Algorithm and Neural Network |
| |
| LI Zhe,HU Li-li. Prediction Model of Ductility of Unbonded Partially Prestressed Concrete Beam with High Strength Based on Genetic Algorithm and Neural Network[J]. Journal of Xi'an University of Technology, 2005, 21(3): 281-284 |
| |
| Authors: | LI Zhe HU Li-li |
| |
| Abstract: | In this paper,the genetic algorithm-neural network model is used to predict the ductility of unbonded partially prestressed concrete beam with high strength.23 beams are selected as learning samples and 3 beams as test samples.The better results are obtained from the model.Elitist strategy and the real number coding are used in the model.Meanwhile,in order to prevent "premature convergence",a new way to change fitness value is suggested. |
| |
| Keywords: | genetic algorithm neural network concrete beam with high strength prestressed ductility |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
