PDF下载 分享
[1]杨晋朝,黄敏超.深度交叉遗传神经网络 在液体火箭发动机故障检测中的应用[J].火箭推进,2009,35(02):41-45.
　Yang Jinzhao,Huang Minchao.Application of deep cross-hybrid genetic neural network to fault detection of liquid rocket engines[J].Journal of Rocket Propulsion,2009,35(02):41-45.
点击复制

深度交叉遗传神经网络 在液体火箭发动机故障检测中的应用

《火箭推进》[ISSN:1672-9374/CN:CN 61-1436/V] 卷: 35 期数: 2009年02期 页码: 41-45 栏目: 测控与试验 出版日期: 2009-04-30
Title:
Application of deep cross-hybrid genetic neural network to fault detection of liquid rocket engines
作者:
杨晋朝; 黄敏超
国防科技大学航天与材料工程学院，湖南长沙410073
Author(s):
Yang Jinzhao; Huang Minchao
Institute of Aerospace and Material Engineering,National University of Defense Technology，Changsha 410073，China
关键词:
遗传算法; BP神经网络。故障检测; 全局优化
Keywords:
genetic algorithm; back propagation neural network; fault detection; global optimization
分类号:
V434
文献标志码:
A
摘要:
将遗传算法与BP神经网络深度交叉融合，即采用遗传算法对BP神经网络的权 值和阈值进行多点优化．而在进化的每一代中随机取少量染色体进行单一BP网络训练．训练 结果再返回染色体，经过若干代的进化后得到稳定的权值和阁值．再将它们赋给BP神经网 络，作为初始值，按误差前向反馈算法沿负梯度搜索重新训练，最终得到最优解。这种算法 既避免BP算法易陷入局部最优解的不足．又克服遗传算法以类似穷举的形式寻找最优解而引 起的搜索时间长、速度慢的缺点。并且经过仿真分析。深度交叉遗传BP神经网络的收敛性和 故障诊断能力优于传统B
Abstract:
This paper proposes a new hybrid algorithm based on genetic algorithm and BP neural network．First，multi—point optimization of the BP neural network's weights and threshold values in GA algorithm is carried out，and some chromosomes that are random sampled in each generation perform single BP neural network training．The result gained above is returned to the chromosomes． Second，stable weights and threshold values are obtained after the evolution of some generations， then they ale used as the initial value to train the BP neural network by seeking along negative grads in error forward feedback algorithm，and finally the global optimum is gained．’11le proposed algorithm can avoid the deficiency of BP algorithm that may easily be steeped in local optimums，and can also overcome GA's shortcomings of long seeking time and low seeking due to the method of enumerating．，11le results of simulation indicates that the ability of convergence and diagnosis of the proposed algorithm is better than that of traditional BP neural network or only using GA，and the algorithm Can be effectively applied to the fault detection of liquid rocket engine．

参考文献/References:

[l]陈明．神经网络模型[M]．大连：大连理工大学出版社，1995．
[2]钟珞，饶文碧，邹承明[M]．人工神经网络及其融合应用技术．北京：科学出版社，2007．
[3]王小平，曹立明·遗传算法——理论、应用与软件实现[M]．西安：西安交通大学出版社，2002·
[4]Yutaka Fukuoka，Hideo Matsuki，Hasuki，Hasruyuki MinamitaIli．et aI．A modified back-propagation method to avoid false local minima[J]．Neural networks，1998，(1 1)：1059_1072．

相似文献/References:

[1]周云端,何志勇,赵瑞国.基于遗传算法的加速度控制PID参数自整定研究[J].火箭推进,2012,38(04):68.
　ZHOU Yun-duan,HE Zhi-yong,ZHAO Rui-guo.Study on genetic algorithm based auto-tuning of acceleration control PID parameters[J].Journal of Rocket Propulsion,2012,38(02):68.
[2]魏鹏飞.一种综合分析LRE减损控制律的 智能方法研究[J].火箭推进,2008,34(03):1.
　Wei Pengfei.Research of an intelligent method to analyze damage-mitigating control law for liquid-propellant rocket engine[J].Journal of Rocket Propulsion,2008,34(02):1.
[3]王鹏,李旭昌,徐颖军,等.固体火箭发动机总体优化设计[J].火箭推进,2007,33(04):16.
　Wang Peng,Li Xuchang,Xu Yingjun.The optimal design of solid rocket motors[J].Journal of Rocket Propulsion,2007,33(02):16.
[4]严俊峰,陈炜.基于遗传算法的低比转速高速泵优化设计[J].火箭推进,2006,32(03):1.
　Yan Junfeng,Chen Wei.Optimum design of low-specific-speed high-speed centrifugal pump based on genetic algorithm[J].Journal of Rocket Propulsion,2006,32(02):1.

备注/Memo

收稿日期：2008—12—26；修回13期：2009-03-03。 作者简介：杨晋朝(1981一)，男，硕士，研究领域为推进系统动力学、控制与健康监控。

更新日期/Last Update: 1900-01-01