首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于BP人工神经网络的人工海水侵蚀对混凝土强度影响的试验研究
引用本文:赵少伟,王铁成,郭磊.基于BP人工神经网络的人工海水侵蚀对混凝土强度影响的试验研究[J].海洋技术,2007,26(4):115-117.
作者姓名:赵少伟  王铁成  郭磊
作者单位:1. 天津大学建筑工程学院,天津,300072;河北工业大学土木工程学院,天津,300132
2. 天津大学建筑工程学院,天津,300072
3. 广东省水利水电科学研究院,广东,广州,510610
基金项目:河北省科技攻关项目(051145)
摘    要:混凝土受盐害侵蚀破坏直接影响混凝土的强度和耐久性。针对混凝土受盐害侵蚀破坏功能函数不能明确表达及非线性程度高的特点,利用BP人工神经网络进行分析,在大量试验数据基础上,通过计算方法的优化和样本的训练,对隐含层和各隐含单元多次试取,最优选取trainglm训练函数,建立盐害预测的人工神经网络系统。解析结果表明,混凝土试件抗压强度预测值和试验实测值的相对误差较小,建立的人工神经网络模型具有较高的预测精度。

关 键 词:人工海水  侵蚀  混凝土  抗压强度  BP网络
文章编号:1003-2029(2007)04-0115-03
收稿时间:2007-07-11
修稿时间:2007年7月11日

Study on Strength of Concrete Corroded by Artificial Sea Water based on BP Artificial Neural Network
ZHAO Shao-wei,WANG Tie-cheng,GUO Lei.Study on Strength of Concrete Corroded by Artificial Sea Water based on BP Artificial Neural Network[J].Ocean Technology,2007,26(4):115-117.
Authors:ZHAO Shao-wei  WANG Tie-cheng  GUO Lei
Abstract:Deteriorated concrete by salt has directly effect on strength and durability. BP-artificial neural network is used to analyze corrosion damage resulted by salt because risk function of deteriorated concrete could not be exactly expressed and possess high nonlinear.A BP- artificial network mode is established on the compressive strength of concrete corroded by salt and optimizes the learning arithmetic and train the net with specimens.The trainglm being best function is chosen by lots of trail to hiding level. The analysis shows that the predictive results are consistent with the experimental results.The artificial network model has high precision.
Keywords:atificial sea  corrosion  concrete  compressive strength  BP-artificial network
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号