首页 | 本学科首页   官方微博 | 高级检索  
     

BP神经网络和RBF神经网络在墩柱抗震性能评估中的比较研究
引用本文:冯清海,袁万城.BP神经网络和RBF神经网络在墩柱抗震性能评估中的比较研究[J].结构工程师,2007,23(5):41-47,69.
作者姓名:冯清海  袁万城
作者单位:同济大学土木工程防灾国家重点实验室,上海,200092
摘    要:通过仿真实例,对BP神经网络和RBF神经网络在墩柱抗震性能评估中的应用进行了比较研究.训练和仿真结果表明,BP神经网络和RBF神经网络均能很好地对墩柱的延性进行抗震性能评估,并且,在相同精度要求下,RBF神经网络比BP神经网络的训练速度更快,效率更高,充分体现了RBF神经网络的优越性.同时也总结了BP神经网络和RBF神经网络所存在的不足.实际应用中,可以以此结论为指导来设计神经网络.

关 键 词:人工智能  BP神经网络  RBF神经网络  性能评估  神经网络  墩柱  抗震  性能评估  比较  研究  Columns  Seismic  Resistance  Performance  Evaluation  RBF  Neural  Network  BP  Neural  Network  设计  指导  存在  效率  训练速度  精度要求  延性  仿真结果  应用
收稿时间:2006-10-10
修稿时间:2006-10-10

Comparative Study on BP Neural Network and RBF Neural Network in Performance Evaluation of Seismic Resistance for Pier Columns
FENG Qinghai,YUAN Wancheng.Comparative Study on BP Neural Network and RBF Neural Network in Performance Evaluation of Seismic Resistance for Pier Columns[J].Structural Engineers,2007,23(5):41-47,69.
Authors:FENG Qinghai  YUAN Wancheng
Affiliation:State Key Laboratory for Disaster Reduction in Civil Engineering, Tongji University, Shanghai 200092, China
Abstract:The BP Network and RBF Network in the use of performance evaluation of seismic resistance for pier columns were compared, and a simulation example was presented. The results of training and simulating indicate that, Both BP Network and RBF Network can evaluate the performance of seismic resistance for pier columns well, and RBF neural network has several advantages over BP neural network on the side of performance evaluation of seismic resistance for pier columns, especially in accuracy and training efficiency. At the same time,some disadvantages of RBF neural network and BP neural network were presented. In practice, the conclusions will provide theoretical guide for the design of neural network.
Keywords:artificial intelligence  BP neural network  RBF neural network  performance evaluation
本文献已被 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