计算机工程与应用 ›› 2008, Vol. 44 ›› Issue (13): 235-237.

• 工程与应用 • 上一篇    下一篇

基于遗传BP算法的神经网络在桩基检测中的应用

戎丽霞1,刘金亮2   

  1. 1.德州学院 计算机系,山东 德州 253023
    2.德州学院 数学系,山东 德州 253023
  • 收稿日期:2007-08-22 修回日期:2007-11-20 出版日期:2008-05-01 发布日期:2008-05-01
  • 通讯作者: 戎丽霞

Application to defect of pile of neural network based on GA-BP algorithm

RONG Li-xia1,LIU Jin-liang2   

  1. 1.Dept. of Computer,Dezhou University,Dezhou,Shandong 253023,China
    2.Dept. of Mathematics,Dezhou University,Dezhou,Shandong 253023,China
  • Received:2007-08-22 Revised:2007-11-20 Online:2008-05-01 Published:2008-05-01
  • Contact: RONG Li-xia

摘要: 文中介绍了传统的遗传算法与BP算法,并分析了它们的不足。提出了一种将遗传算法和BP算法相结合的遗传BP算法,在遗传算法的群体设计中采用小生境技术,防止优秀个体早期退化。并将此方法应用于桩基检测中,实验证明了其正确性和有效性。

关键词: BP算法, 遗传算法, 小生境技术, 神经网络, 桩基检测

Abstract: In this paper,the principle and characteristics of Genetic Algorithm and BP algorithm are introduced.Their disadvantages are analyzed.A new algorithm on combination of genetic algorithm and back propagation algorithm is proposed.In genetic algorithm,niche technology is applied in order to avoid the early degenerate of good network.The experiment in defect of pile illustrates that the new algorithm has better efficiency.

Key words: BP algorithm, genetic algorithm, niche technique, neural network, defect of pile