首页 > 论文 > 电光与控制 > 25卷 > 3期(pp:68-72)

LSTM网络和ARMA模型对惯性器件随机误差预测适应性分析

Adaptability of LSTM Network and ARMA Modeling to Random Error Prediction of Inertial Devices

  • 摘要
  • 论文信息
  • 参考文献
  • 被引情况
  • PDF全文
分享:

摘要

结合惯性器件随机误差研究的实际, 针对传统的基于时间序列的ARMA建模方法和深度学习LSTM网络进行了适用性和实时性对比分析, 通过获取具体型号的惯性器件输出数据设计了算例。研究认为在未做实时性要求的情况下, ARMA建模和LSTM网络均可以达到较好的拟合效果, 而建立LSTM网络方法可以减少提取趋势项和周期项的环节; 实时在线预测情况下LSTM网络优势明显, 但预测精度会随时间序列缩短而明显下降, 可以在一定程度上反映噪声变化的趋势并据此对整体控制系统的滤波算法进行优化。

Abstract

For the random error of inertial devicesa contrastive analysis was made to the applicability and real-time performance of the traditional ARMA modeling method and the popular deep learning LSTM network.A simulation example was designed by aobtaining the output data of a specific inertial device.The study showed that:1) Without the real-time performance requirementboth ARMA modeling and LSTM network can achieve accurate prediction resultsand the LSTM network is advantageous since it doesnt need to extract the signal trend and cycle terms;and 2) In real-time predictionthe LSTM network has obvious advantagesbut the prediction accuracy decreases with the shortening of time serieswhereas the output can still reflect the trend of noise change and can be used for optimizing the filtering algorithm of the whole control system.

广告组1 - 空间光调制器+DMD
补充资料

中图分类号:V448

DOI:10.3969/j.issn.1671-637x.2018.03.015

所属栏目:工程应用

收稿日期:2018-01-03

修改稿日期:2018-01-25

网络出版日期:--

作者单位    点击查看

杨 其:火箭军工程大学,a.士官学院, 山东 青州 262500b.初级指挥学院, 西安 710025
陈水忠:中国航空工业集团公司洛阳电光设备研究所, 河南 洛阳 471000
沈淑梅:中国航空工业集团公司洛阳电光设备研究所, 河南 洛阳 471000
朱振华:中国航空工业集团公司洛阳电光设备研究所, 河南 洛阳 471000

联系人作者:联系作者

备注:杨 其(1983 —),男, 山东菏泽人, 讲师, 博士生, 研究方向为飞行器动力学与制导。

【1】RAU′L D CLUIS CJESU′S S.Guidance and control for high dynamic rotating artillery rockets[J].Aerospace Science and Technology201764(5):204-212.

【2】MICHAEL ABLENDERHEIDI W.Flight dynamics of a hypersonic vehicle during inlet un-start[C]//The 16th AIAA/DLR/DGLR International Space Planes and Hypersonic Systems and Technologies Conference2009:AIAA-2009-7292.

【3】BO WZHI H DCHENG Let al.Estimation of information sharing error by dynamic deformation between inertial navigation system [J].IEEE Transactions on Industrial Electronics201461(4):2015-2023.

【4】SAVAGE P G.Analytical modeling of sensor quantization in strapdown inertial navigation error equations[J].Journal of GuidanceControl and Dynamics200225(5):833-842.

【5】张瑞民, 杨其, 魏诗卉, 等.捷联惯测组合标定及误差补偿技术[M].北京:国防工业出版社, 2017.

【6】STHAL FDEVEL MIMBAUD Jet al.Study on the origin of 1/f noise in quartz resonators[J].Journal of Statistical Mechanics Theory & Experiment2016(5):054025.

【7】JONES B K.Electrical noise as a measure of quality and reliability in electronic devices.[J] Advances in Electronics and Electron Physics199387(1):201-257.

【8】BROWN R GHWANG P Y C.Introduction to random signals and applied Kalman filtering[M].3rd ed.New York:John Wiley & SonsInc.1997.

【9】韩松来.GPS和捷联惯导组合导航新方法及系统误差补偿方案研究[D].长沙:国防科学技术大学2010.

【10】HOCHREITER SSCHMIDHUBER.J.Long short-term memory[J].Neural Computation19979(8):1735-1780.

【11】GERS F ASCHMIDHUBER JCUMMINS F.Learning to forget:continual prediction with LSTM[J].Neural Computation200012(10):2451-2471.

【12】HOCHREITER SBENGIO YFRASCONI P.Gradient flow in recurrent nets:the difficulty of learning long-term dependencies[J].Field Guide to Dynamical Recurrent Networks200128(2):237-243.

【13】BENGIO SBENGIO Y.Taking on the curse of dimensionality in joint distributions using neural networks[J].IEEE Transactions on Neural NetworksSpecial Issue on Data Mining and Knowledge Discovery200011(3):550-557.

【14】SUTSKEVER IVINYALS OLE Q V.Sequence to sequence learning with neural networks[J].Computer Science, 2014(4):310-311.

【15】LIPTON Z CBERKOWITZ J.A critical review of recurrent neural networks for sequence learning[J].Computer Science, 2015(1):104-112.

【16】HASIM SANDREW SFRANCOISE B.Long short-term memory current neural networks architecture for large scale acoustic modeling[C]//Proceeding of Interspeech ISCA2014:338-342.

【17】GRAVE A.Generating sequences with recurrent neural networks[J].Computer Science2013(1):234-243.

【18】PALANGI HDENG LSHEN Yet al.Deep sentence embedding using long short-term memory networks:analysis and application to information retrieval[J].IEEE/ACM Transactions on Audio Speech & Language Processing 201524(4):694-707.

引用该论文

YANG Qi,CHEN Shui-zhong,SHEN Shu-mei,ZHU Zhen-hua. Adaptability of LSTM Network and ARMA Modeling to Random Error Prediction of Inertial Devices[J]. Electronics Optics & Control, 2018, 25(3): 68-72

杨 其,陈水忠,沈淑梅,朱振华. LSTM网络和ARMA模型对惯性器件随机误差预测适应性分析[J]. 电光与控制, 2018, 25(3): 68-72

您的浏览器不支持PDF插件,请使用最新的(Chrome/Fire Fox等)浏览器.或者您还可以点击此处下载该论文PDF