|本期目录/Table of Contents|

[1]王彩霞,刘义艳.基于VMD和DBN的结构健康状态趋势预测[J].工业仪表与自动化装置,2019,(06):24-29.[doi:1000-0682(2019)06-0000-00]
　WANG Caixai,LIU Yiyan.Trend prediction of structural health based on VMD and DBN[J].Industrial Instrumentation & Automation,2019,(06):24-29.[doi:1000-0682(2019)06-0000-00]
点击复制

基于VMD和DBN的结构健康状态趋势预测

《工业仪表与自动化装置》[ISSN:1000-0682/CN:61-1121/TH]

卷:
期数:
2019年06期
页码:
24-29
栏目:
出版日期:
2019-12-15

文章信息/Info

Title:
Trend prediction of structural health based on VMD and DBN
作者:
王彩霞; 刘义艳
长安大学 电子与控制工程学院，西安 710064
Author(s):
WANG Caixai;  LIU Yiyan
School of Electronic and Control Engineering, Chang’an University, Xi’an 710064, China
关键词:
变分模态分解; 深度置信网络; 结构健康预测; 动量学习率
Keywords:
variational modal decomposition;  deep belief network;  structural health prediction;  momentum learning rate
分类号:
TU317
DOI:
1000-0682(2019)06-0000-00
文献标志码:
A
摘要:
为了能更好地反映结构状态变化趋势，提高限制玻尔兹曼机(RBM)对非线性信号的特征提取能力，该文在深度置信网络(DBN)的预训练阶段引入动量学习率，提出了一种基于VMD-DBN的结构健康状态趋势预测方法。将待处理信号用变分模态分解(VMD)方法进行分解，对分解的分量经希尔伯特变换得到瞬时频率；将瞬时频率作为改进DBN预测模型的输入进行结构健康状态趋势预测。仿真和工程实验表明，VMD方法有效地消除了模态混叠，分解出信号的各个固有分量。改进DBN模型的预测精度优于传统DBN、长短时记忆网络(LSTM)和传统BP神经网络，说明改进DBN模型能够很好地应用于结构健康状态趋势预测。
Abstract:
In order to better reflect the trend of structural state change, the limited Boltzmann machine(RBM) can improve the feature extraction ability of nonlinear signals.The momentum learning rate is introduced in the pre-training stage of the deep confidence network(DBN), and a structural health state trend prediction method based on VMD-DBN is proposed.The signal to be processed is decomposed by the variational modal decomposition(VMD) method, and the decomposed component is transformed by Hilbert to obtain the instantaneous frequency.The instantaneous frequency is used as the input of the improved DBN prediction model to predict the structural health state trend. Simulation and engineering experiments show that the VMD method can eliminate mode aliasing and decompose each intrinsic component of the signal. The prediction accuracy of the improved DBN model is better than that of the traditional DBN model, the LSTM model and the traditional BP neural network.The improved DBN model can be well applied to the prediction of structural health trend.

参考文献/References:

[1] Paulraj M P,Yaacob S,Majid M S A,et al.Structural Steel Plate Damage Detection using Non Destructi-ve Testing, Frame Energy Based Statistical Features and Artificial Neural Networks[J].Procedia Engineering,2013,53(0): 376-386.

[2] 刘茹.基于FESN的结构健康状态智能预测研究[D].西安:长安大学,2017.
[3] 李鹏.基于新形势下的数字信号处理技术的应用和发展研究[J].电子测试,2016(6):76-77.
[4] 朱炳安.基于小波分析对变速箱的故障诊断[J].机械设计与制造,2016(11):245-248.
[5] 吴超,赵军,郭天太,等.EMD与JADE相结合的算法及验证[J].中国计量学院学报,2015,26(3):366-372.
[6] Dragomiretskiy K.Zosso D.Variational mode decomposition [J].IEEE Transactions on Signal Processing,2014,62(3): 531-544.
[7] Dedinec A,Filiposka S,Dedinec A,et al.Deep belief network based electricity load forecasting: An analys is of Macedonian case[J]. Energy, 2016,115:1688-1700.
[8] 吴坚,郑照红,薛家祥.深度置信网络光伏发电短时功率预测研究[J].中国测试,2018,v.44,No.238(05):14-19.
[9] 邵双双,刘丽冰,谭志洪.改进的深度信念网络预测模型及其应用[J].计算机应用, 2018(A01):28-31.
[10] Fang Wang, Suxia Ma etal.Prediction of NOx emission for coal-fired boilers based on deep belief network[J].Control Engineering Practice,2018:26-35.
[11] 孟凡煦,张琰骏.基于VMD和ELM的超短期负荷预测方法[J].国外电子测量技术,2019(1):24-28.
[12] Hinton G E,Osindero S,TehYW.A fast learning algorithm for deep belief nets[J].Neural Computation,2006,1(7): 1527-1554.
[13] Nair V,Hinton G E.Rectified linear units improve restricted Boltzmann machines[C]//27th International Conference on Machine Learning.Haifa,Israel:International Machine Learning Society,2010:807-814．
[14] Hinton GE.A practical guide to training restricted boltzmann machines[J].Momentum,2012,9(1):599-619.
[15] 刘义艳.结构健康监测与智能诊断技术研究[D].西安:长安大学,2010.[16] http://cive.seas.wustl.edu/wusceel/ASCE.shm/experim_ phase _2.htm.

相似文献/References:

[1]孙 涛,朱 愈,胡振中.基于PLC的斗轮机无人值守智能控制策略设计[J].工业仪表与自动化装置,2022,(06):100.[doi:10.19950/j.cnki.cn61-1121/th.2022.06.019]
　SUN Tao,ZHU Yu,HU Zhenzhong.Design of unattended intelligent control strategy for bucket wheel machine based on PLC[J].Industrial Instrumentation & Automation,2022,(06):100.[doi:10.19950/j.cnki.cn61-1121/th.2022.06.019]

备注/Memo

备注/Memo:
收稿日期：2019-04-08
基金项目：国家自然科学基金青年基金(61701044)
作者简介：王彩霞(1990)，女，硕士研究生，研究方向为结构健康监测与损伤诊断。
更新日期/Last Update: 1900-01-01