基于优化概率神经网络的飞机结冰情况识别

丁娣; 钱炜祺; 汪清

doi:10.7638/kqdlxxb-2021.0022

欢迎来到能源电力期刊网！

English Version 期刊登录

您当前的位置：

首页 >

文章列表页 >

基于优化概率神经网络的飞机结冰情况识别

研究论文 | 更新时间：2026-05-28

- 基于优化概率神经网络的飞机结冰情况识别
- Aircraft icing classification using optimized probabilistic neural networks
- 空气动力学学报 2022年40卷第5期页码：100-109
- 作者机构：
  
  1. 中国空气动力研究与发展中心空气动力学国家重点实验室,绵阳,621000
  2. 中国空气动力研究与发展中心计算空气动力研究所,绵阳,621000
- 作者简介：
  
  丁娣*（1981-），女，博士，副研究员，研究方向：飞行力学，参数辨识，飞机结冰探测. E-mail：dingdi1981@hotmail.com
- 基金信息：
  
  国家自然科学基金（11802325）
- DOI：10.7638/kqdlxxb-2021.0022
  中图分类号： TP183
- 纸质出版：2022
- 稿件说明：
移动端阅览
丁娣, 钱炜祺, 汪清. 基于优化概率神经网络的飞机结冰情况识别[J]. 空气动力学学报, 2022, 40(5): 100−109.

DING Di, QIAN Weiqi, WANG Qing. Aircraft icing classification using optimized probabilistic neural networks[J]. 2022, 40(5): 100-109.
丁娣, 钱炜祺, 汪清. 基于优化概率神经网络的飞机结冰情况识别[J]. 空气动力学学报, 2022, 40(5): 100−109. DOI： 10.7638/kqdlxxb-2021.0022

DING Di, QIAN Weiqi, WANG Qing. Aircraft icing classification using optimized probabilistic neural networks[J]. 2022, 40(5): 100-109. DOI： 10.7638/kqdlxxb-2021.0022.

摘要

针对飞机飞行过程结冰情况识别和预警需求，利用概率神经网络开展直接基于飞行状态的飞机结冰情况识别研究。首先利用双水獭结冰研究飞机的结冰数据，建立结冰飞行动力学模型，生成飞机在干净外形、中度和重度机翼结冰以及中度和重度尾翼结冰五种情况下的大量飞行仿真数据，作为分类神经网络训练和性能评估的基础；再基于大量飞行仿真数据分析各飞行状态变量受结冰影响的显著程度，选择受结冰影响较大的6个状态变量作为识别网络的输入；提出了一种传播参数优化策略对基于概率神经网络的结冰分类精度进行优化，在此基础上对6个状态输入的结冰分类结果展开对比分析，发现基于这6个受结冰影响较大的状态变量构建的优化概率神经网络，对于训练和评估数据均具有较高的分类精度，对于训练数据分类精度均可达到100%，对于评估数据，分类精度最高的网络正确率可达99.1%，效果最差的网络分类精度也超过85%；快变量对于评估数据分类精度较高，但对泛化性评估数据的分类精度偏低，慢变量则正好相反；基于快变量攻角

和慢变量位移

的神经网络分类效果相对较好，若不考虑风场扰动的变化，基于攻角

的分类网络的最大误报概率不超过1.4%，若综合考虑风场扰动变化的情况，则采用基于

的分类网络效果更好，对于五种结冰情况最大误报概率整体不超过10%。最后利用两种支持向量机方法与优化概率神经网络的分类精度进行对比，进一步验证该方法的性能。

Abstract

This paper issues an aircraft icing classification problem by directly using the flight state variables and probabilistic neural networks (PNN) for meeting the needs of classifying and alarming icing situation during aircraft flight. A dynamic model for inflight icing is presented based on the icing experiment research data of the NASA Twin Otter airplane. The dynamic model sim

ulates five situations

i.e.

clean

moderate/severe wing icing

moderate/severe tailplane icing

to generate abundant flight data for the networks’ training and assessment. The influence of icing on different flight state variables is quantitatively analyzed

and the six variables significantly affected by icing are chosen for building the neural networks. A scheme for optimizing the network propagation parameter is presented to improve the classification accuracy

then the performance of the six state variables-based classification networks is analyzed and compared. The results indicate that PNNs built by the six state variables have relatively high accuracy for both the training and validation data. For the training data

the classification accuracy all reaches 100%

while for the validation data

the highest accuracy among all six nets is 99.1%

and the lowest accuracy is still above 85%. Fast variables have a higher accuracy for the validation data than slow variables

but a lower accuracy for the generalization assessment data

whereas the slow variables conduct oppositely. The networks built by the fast variable

angle of attack

and the slow variable

displacement

perform relatively better than the others. Without the wind disturbance

the most considerable false alarming probability of

-based networks is below 1.4%. With the wind disturbance

the

-based networks are more preferable

as the overall false alarming probability for all the five icing situations is below 10%. Finally

the performance of the optimized PNN is further validated by comparing the classification accuracy with that of two support vector machine methods.

关键词

Keywords

references

CAO Y H, TAN W Y, WU Z L. Aircraft icing: an ongoing threat to aviation safety[J]. Aerospace Science and Technology, 2018, 75: 353-385. DOI: 10.1016/j.ast.2017.12.028

CAO Y H, WU Z L, SU Y, et al. Aircraft flight characteristics in icing conditions[J]. Progress in Aerospace Sciences, 2015, 74: 62-80. DOI: 10.1016/j.paerosci.2014.12.001

GORAJ Z. An overview of the de-icing and anti-icing technologies with prospects for the future[C]//24th International Congress of Aeronautical Sciences, Yokohama, Japan, 2004.

JARVINEN P. aircraft ice detection method[C ] //45th AIAA Aerospace Sciences Meeting and Exhibit, Reno, Nevada. Reston, Virginia: AIAA, 2007. doi: 10.2514/6.2007-696

HUANG X, TEPYLO N, POMMIER-BUDINGER V, et al. A survey of icephobic coatings and their potential use in a hybrid coating/active ice protection system for aerospace applications[J]. Progress in Aerospace Sciences, 2019, 105: 74-97. DOI: 10.1016/j.paerosci.2019.01.002

SHARIFI N, DOLATABADI A, PUGH M, et al. Anti-icing performance and durability of suspension plasma sprayed TiO 2 coatings[J ] . Cold Regions Science and Technology, 2019, 159: 1-12. DOI: 10.1016/j.coldregions.2018.11.018

LIU J P, WANG J, MEMON H, et al. Hydrophobic/icephobic coatings based on thermal sprayed metallic layers with subsequent surface functionalization[J]. Surface and Coatings Technology, 2019, 357: 267-272. DOI: 10.1016/j.surfcoat.2018.10.002

SCHUCHARD E, MELODY J, BASAR T, et al. Detection and classification of aircraft icing using neural networks[C ] // 38th Aerospace Sciences Meeting and Exhibit, Reno, NV. Reston, Virginia: AIAA, 2000. doi: 10.2514/6.2000-361

DING D, CHE J, QIAN W Q. Fault-tolerant flight control for aircraft wing icing based on icing detection method[C ] // 2018 IEEE CSAA Guidance, Navigation and Control Conference (CGNCC), Xiamen, China. IEEE, 2018: 1-6. doi: 10.1109/GNCC42960.2018.9018871

WEI Y, XU H J, XUE Y. Adaptive neural networks-based dynamic inversion applied to reconfigurable flight control and envelope protection under icing conditions[J]. IEEE Access, 2020, 8: 11577-11594. DOI: 10.1109/ACCESS.2020.2964728

MCCANN D W. NNICE - a neural network aircraft icing algorithm[J]. Environmental Modelling Software, 2005, 20(10): 1335-1342. DOI: 10.1016/j.envsoft.2004.09.027

CAO Y H, YUAN K G, LI G Z. Effects of ice geometry on airfoil performance using neural networks prediction[J]. Aircraft Engineering and Aerospace Technology, 2011, 83(5): 266-274. DOI: 10.1108/00022661111159870

OGRETIM E, HUEBSCH W, SHINN A. Aircraft ice accretion prediction based on neural networks[J]. Journal of Aircraft, 2006, 43(1): 233-240. DOI: 10.2514/1.16241

CHANG S N, LENG M Y, WU H W, et al. Aircraft ice accretion prediction using neural network and wavelet packet transform[J]. Aircraft Engineering and Aerospace Technology, 2016, 88(1): 128-136. DOI: 10.1108/aeat-05-2014-0057

CALISKAN F. Neural network based icing identification and fault tolerant control of a 340 aircraft[J]. International Journal of Electrical, Computer, and Systems Engineering, 2007, 1(2): 98–103.

AYKAN R, CALISKAN F, HAJIYEV C. F16 icing identification based on neural networks[J]. IFAC Proceedings Volumes, 2004, 37(19): 201-206. DOI: 10.1016/S1474-6670(17)30683-3

MELODY J, POKHARIYAL D, MERRET J, et al. Sensor integration for inflight icing characterization using neural networks[C ] // 39th Aerospace Sciences Meeting and Exhibit, Reno, NV. Reston, Virginia: AIAA, 2001. doi: 10.2514/6.2001-542

DONG Y Q. An application of Deep Neural Networks to the in-flight parameter identification for detection and characterization of aircraft icing[J]. Aerospace Science and Technology, 2018, 77: 34-49. DOI: 10.1016/j.ast.2018.02.026

DONG Y Q, AI J L. Research on inflight parameter identification and icing location detection of the aircraft[J]. Aerospace Science and Technology, 2013, 29(1): 305-312. DOI: 10.1016/j.ast.2013.03.012

DONG Y Q, AI J L. Inflight parameter identification and icing location detection of the aircraft: the time-varying case[J]. Journal of Control Science and Engineering, 2014, 2014: 1-11. DOI: 10.1155/2014/396532

JOHNSON M D, ROKHSAZ K. Using artificial neural networks and self-organizing maps for detection of airframe icing[J]. Journal of Aircraft, 2001, 38(2): 224-230. DOI: 10.2514/2.2779

YING S B, AI J L. Detection of aircraft in-flight icing in non-steady atmosphere using artificial neural network[C ] // 2010 Third International Symposium on Intelligent Information Technology and Security Informatics, Jian, China. IEEE, 2010: 206-211. doi: 10.1109/IITSI.2010.146

SPECHT D F. Probabilistic neural networks[J]. Neural Networks, 1990, 3(1): 109-118. DOI: 10.1016/0893-6080(90)90049-Q

SPECHT. Probabilistic neural networks for classification, mapping, or associative memory[C ] // IEEE 1988 International Conference on Neural Networks, San Diego, CA, USA. IEEE, 1988: 525-532. DOI: 10.1109/ICNN.1988.23887

陈明等. MATLAB神经网络原理与实例精解[M]. 北京: 清华大学出版社, 2013: 205-207

MELODY J W. Inflight characterization of aircraft icing[D]. Illinois Urbana: University of Illinois at Urbana-Champaign Graduate College, 2004: 17-46.

MELODY J W, HILLBRAND T, BAŞAR T, et al. H∞ parameter identification for inflight detection of aircraft icing: the time-varying case[J]. Control Engineering Practice, 2001, 9(12): 1327-1335. DOI: 10.1016/S0967-0661(01)00081-8

ZHAO Y K, WANG Y L. Research on anti-disturbance for quadrotor aircraft in wind field[J]. Mechanical Science and Technology for Aerospace Engineering, 2019, 38(4): 530-537. (in Chinese)

KUSY M, ZAJDEL R. Probabilistic neural network training procedure based on Q(0)-learning algorithm in medical data classification[J]. Applied Intelligence, 2014, 41(3): 837-854. DOI: 10.1007/s10489-014-0562-9

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

基于Ac-准则的最优输入设计方法

风扇叶片颗粒侵蚀特性的多因素分析与模拟

离心通风机降噪研究进展

基于LBM方法的LAGOON起落架噪声预测与降噪方法研究

安装效应对螺旋桨纯音噪声的影响