基于遗传BP算法的神经网络在桩基检测中的应用  

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

作　　者：戎丽霞[1] 刘金亮[2] 

机构地区：[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

出　　处：《计算机工程与应用》2008年第13期235-237,共3页Computer Engineering and Applications

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

关 键 词：BP算法 遗传算法 小生境技术 神经网络 桩基检测 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]