Risk assessment method for civil aircraft approach and landing at high plateau based on QAR data
-
摘要:
民机高高原进近着陆是高原飞行高风险阶段。为有效实施高高原进近着陆风险识别和等级判据,提出基于熵权可变模糊识别的长短时记忆网络与深度神经网络(LSTM-DNN)相融合的深度学习风险评估方法。基于快速存取记录器(QAR)记录的高高原飞行数据,借鉴民机飞行品质监控(FOQA)咨询通告和行业QAR 监控标准,结合指标重要度分析与Delphi专家调查,提取着陆时航向变化大、航迹低、610~305 m进近时下降率大、接地时垂直加速度及153~15 m进近时下降率大5个关键监控项目作为民机高高原进近着陆风险评估指标。为克服评估指标权重主观性偏差,应用熵权法确定评估指标权重,基于可变模糊识别方法构建风险等级隶属函数,建立基于LSTM-DNN的民机高高原进近着陆风险评估模型。以成都—拉萨进近着陆航段为例,提取QAR数据,对该风险评估模型进行训练与测试,并与Logistic多元回归、支持向量机(SVM)等评估方法进行比较,结果表明:所提方法平均准确率达到94.18%,最高可达94.79%,验证了方法的客观有效性。
Abstract:The high plateau approach and landing of civil aircraft is a high-risk stage of high plateau flight. To effectively identify the risk and its grade of this approach and landing, a long short term memory-deep neural network (LSTM-DNN) deep learning risk assessment method is proposed based on the variable fuzzy identification of entropy weights. This method utilizes high-altitude flight data recorded by the quick access recorder (QAR), referencing the advisory notices from the flight operations quality assurance (FOQA) of civil aviation as well as the industry QAR monitoring standards. The method combines indicator importance analysis with Delphi expert surveys to extract five key monitoring items for civil aviation high-altitude approach and landing risk assessment, including significant changes in heading during landing, low trajectory, large descent rate during the 610−305 m approach, touchdown vertical acceleration during landing, and high descent rate during the 153−15 m approach. To overcome the subjective bias of the evaluation index weight, the entropy weight method is then used to determine the evaluation index weight, with the risk level membership function constructed based on the variable fuzzy identification method. Finally, the LSTM-DNN risk assessment model for civil aircraft approach and landing at high plateau is established. Taking the Chengdu−Lhasa approach and landing segment as an example, this study extracted the QAR data to train and test the risk assessment model, and compared the results with those of the evaluation methods such as Logistic multiple regression and support vector machines (SVM). The results show that the recognition rate of the proposed method reaches 94.18% on average with the highest being 94.79%, verifying the effectiveness of the method.
-
表 1 空客系列飞机飞FOQA监控项目规范指标(部分)[14]
Table 1. Specification indicators of FOQA project for airbus series aircraft (partial)[14]
监控项目 监控参数 监控点
下降率大
IVV,ALL610~305 m(含) 305~152 m(含) 152~15 m(含)
进近滚转角大
滚转角,ALL457~152 m(含) 152~61 m(含) 61~15 m(含) 着陆滚转角大 滚转角,ALL 15 m至所有机轮接地 低高速使用减速板 减速板,ALL 使用减速板 进近速度小 空速,ALL 305~15 m(含) 进近速度大 空速,ALL 152~15 m(含) 着陆速度大 空速,ALL 15 m以下 ILS下滑道偏离 下滑道偏离,ALL 305 m以下 ILS航向道偏离 航向道偏离,ALL 305 m以下 低空大速度 空速,ALL 762 m以下 表 2 FOQA风险评价指标分级标准
Table 2. Grading standard of FOQA risk evaluation index
风险等级 着陆时航向
变化大/(°)153~15 m进近时
下降率大/(m·min−1)航迹低/(°) 610~305 m进近时
下降率大/(m·min−1)接地时垂直
加速度/g0级 <4 ≥−700 ≤2.2 ≥−1200 ≤1.5 Ⅰ级 [4,5] [−800,−700) (2.2,2.4] [−1500,−1200) (1.5,1.6] Ⅱ级 (5,6) (−900,−800) (2.4,2.6) (−1800,−1500) (1.6,1.75) Ⅲ级 ≥6 ≤−900 ≥2.6 ≤−1800 ≥1.75 表 3 4种参数组合下综合相对隶属度
Table 3. Comprehensive relative membership degree under combination of four parameters
参数组合 综合相对隶属度 H 0级 Ⅰ 级 Ⅱ 级 Ⅲ 级 a=1
p=10.8935 0.1065 0 0 1.1065 a=1
p=20.7680 0.2320 0 0 1.2320 a=2
p=10.7207 0.2793 0 0 1.2793 a=2
p=20.9385 0.0614 0 0 1.0614 表 4 模糊评价识别结果
Table 4. Fuzzy evaluation and recognition results
样本组 着陆时航向
变化大/(°)153~15 m进近时
下降率大/(m·min−1)航迹低/(°) 610~305 m进近时
下降率大/(m·min−1)接地时垂直
加速度/g等级 1 1.8 −800 0.01 −912 0.96 0级 2 1.8 −832 0.08 −976 1.0 0级 3 4 −832 5 −1316 1.15 Ⅰ级 4 9 −848 7 −1558 0.5 Ⅱ级 5 0.6 −864 0.1 −800 0.98 Ⅰ级 499 2 −656 0 −999 2.80 Ⅱ级 500 8 −672 5 −988 3.78 Ⅲ级 表 5 不同隐藏层DNN预测值
Table 5. Predicted values of DNN with different hidden layers
隐藏层数 预测值 1 0.00701 3 0.00201 5 0.00794 表 6 成都—拉萨进近着陆航段QAR数据(部分)
Table 6. QAR data of Chengdu−Lhasa approach and landing segment (partial)
样本数量 着陆时航向
变化大/(°)153~15 m进近时
下降率大/(m·min−1)航迹低/(°) 610~305 m进近时
下降率大/(m·min−1)接地时垂直
加速度/g等级 1 0.9 −763 0.11 −896 0.99 0 级 2 0.6 −729 0.11 −880 1.01 0 级 50 1.6 −700 8 −832 0.96 0 级 51 2 −649 4 −848 1.01 Ⅰ 级 99 1 −832 4 −730 2.44 Ⅱ 级 100 1.7 −630 −0.67 −726 1.01 0 级 -
[1] 中国民用航空局飞行标准司. 高原机场运行: AC-121-FS-2015-21R1[S]. 北京: 中国民用航空局, 2015.Flight Standards Department of Civil Aviation Administration of China. Plateau airport operation: AC-121-FS-2015-21R1[S]. Beijing: Civil Aviation Administration of China, 2015(in Chinese). [2] 杨志刚, 张炯, 李博, 等. 民用飞机智能飞行技术综述[J]. 航空学报, 2021, 42(4): 525198.YANG Z G, ZHANG J, LI B, et al. Reviews on intelligent flight technology of civil aircraft[J]. Acta Aeronautica et Astronautica Sinica, 2021, 42(4): 525198(in Chinese). [3] 罗婷婷, 孙瑞山. 基于Petri网的进近操作管理模型研究[J]. 中国安全生产科学技术, 2013, 9(11): 137-141.LUO T T, SUN R S. Study on management system model for approach based on Petri net[J]. Journal of Safety Science and Technology, 2013, 9(11): 137-141(in Chinese). [4] International Civil Aviation Organization. Safety management manual: Doc 9859[S]. Montreal: International Civil Aviation Organization, 2018. [5] XU Y J, HOU W K, LI W Z, et al. Aero-engine gas-path fault diagnosis based on spatial structural characteristics of QAR data[C]//Proceedings of the Annual Reliability and Maintainability Symposium. Piscataway: IEEE Press, 2018: 1-7. [6] KLUEVER C A. Unpowered approach and landing guidance with normal acceleration limitations[J]. Journal of Guidance, Control, and Dynamics, 2007, 30(3): 882-885. doi: 10.2514/1.28081 [7] 周长春, 胡栋栋. 基于灰色聚类方法的航空公司飞机进近着陆阶段安全性评估[J]. 中国安全生产科学技术, 2012, 8(7): 99-102.ZHOU C C, HU D D. Safety assessment of aircraft during approach-landing stage based on grey clustering method[J]. Journal of Safety Science and Technology, 2012, 8(7): 99-102(in Chinese). [8] 高小霞, 霍纬纲, 冯兴杰. 基于模糊关联分类器的民机超限事件诊断方法[J]. 北京航空航天大学学报, 2014, 40(10): 1366-1371.GAO X X, HUO W G, FENG X J. Civil aircraft’s exceedance event diagnosis method based on fuzzy associative classifier[J]. Journal of Beijing University of Aeronautics and Astronautics, 2014, 40(10): 1366-1371(in Chinese). [9] 祁明亮, 邵雪焱, 池宏. QAR超限事件飞行操作风险诊断方法[J]. 北京航空航天大学学报, 2011, 37(10): 1207-1210.QI M L, SHAO X Y, CHI H. Flight operations risk diagnosis method on quick-access-record exceedance[J]. Journal of Beijing University of Aeronautics and Astronautics, 2011, 37(10): 1207-1210(in Chinese). [10] WANG L, ZHANG J Y, DONG C T, et al. A method of applying flight data to evaluate landing operation performance[J]. Ergonomics, 2019, 62(2): 171-180. doi: 10.1080/00140139.2018.1502806 [11] 鲁志东, 张曙光, 戴闰志, 等. 大型民机进近着陆段异常能量风险判据[J]. 航空学报, 2021, 42(6): 624132.LU Z D, ZHANG S G, DAI R Z, et al. Abnormal energy risk criteria of large civil airplanes in approach and landing[J]. Acta Aeronautica et Astronautica Sinica, 2021, 42(6): 624132(in Chinese). [12] 李哲, 徐浩军, 薛源, 等. 基于风险预测的飞行安全操纵空间构建方法[J]. 北京航空航天大学学报, 2018, 44(9): 1839-1849.LI Z, XU H J, XUE Y, et al. Construction method of flight safety manipulation space based on risk prediction[J]. Journal of Beijing University of Aeronautics and Astronautics, 2018, 44(9): 1839-1849(in Chinese). [13] 徐骏驰, 孙有朝. 运输类飞机高高原机场运行的适航验证方法[J]. 航空计算技术, 2018, 48(2): 129-134.XU J C, SUN Y C. Airworthiness requirement of transportation category aircraft operation on high altitude airports[J]. Aeronautical Computing Technique, 2018, 48(2): 129-134(in Chinese). [14] 中国民用航空局飞行标准司. 飞行品质监控(FOQA)实施与管理: AC-121/135-FS-2012-45R1[S]. 北京: 中国民用航空局, 2012.Flight Standards Department of Civil Aviation Administration of China. Implementation and management of flight operation quality assurance(FOQA): AC-121/135-FS-2012-45R1[S]. Beijing: Civil Aviation Administration of China, 2012(in Chinese). [15] CUI L F, ZHANG C Q, ZHANG Q Z, et al. A method for aero-engine gas path anomaly detection based on Markov transition field and multi-LSTM[J]. Aerospace, 2021, 8(12): 374. doi: 10.3390/aerospace8120374 [16] ATASOY V E, CETEK C. Enhanced cruise range prediction for narrow-body turbofan commercial aircraft based on QAR data[J]. The Aeronautical Journal, 2021, 125(1286): 672-701. doi: 10.1017/aer.2020.121 [17] 张鹏, 杨涛, 刘亚楠, 等. 基于CNN-LSTM的QAR数据特征提取与预测[J]. 计算机应用研究, 2019, 36(10): 2958-2961.ZHANG P, YANG T, LIU Y N, et al. Feature extraction and prediction of QAR data based on CNN-LSTM[J]. Application Research of Computers, 2019, 36(10): 2958-2961(in Chinese). [18] 刘柳. 基于QAR数据的着陆阶段飞行风险研究[D]. 重庆: 重庆大学, 2018.LIU L. Research on flight risk in landing using QAR data[D]. Chongqing: Chongqing University, 2018(in Chinese). [19] 郑磊, 池宏, 邵雪焱. 基于QAR数据的飞行操作模式及其风险分析[J]. 中国管理科学, 2017, 25(10): 109-118.ZHENG L, CHI H, SHAO X Y. Pattern recognition and risk analysis for flight operations[J]. Chinese Journal of Management Science, 2017, 25(10): 109-118(in Chinese). [20] 李俊辰. 基于机器学习的起飞、进近阶段飞行成绩评估[D]. 天津: 中国民航大学, 2020.LI J C. Flight performance evaluation during takeoff and approach based on machinelearning[D]. Tianjin: Civil Aviation University of China, 2020(in Chinese). [21] KONDO S, KIDERA S, KIRIMOTO T. Target detection algorithm using independent component analysis for pulse Doppler radar[J]. IEICE Communications Express, 2014, 3(7): 211-216. doi: 10.1587/comex.3.211 [22] 陈勇刚, 熊升华, 贺强, 等. 通用航空机队设备可靠性动态识别模型[J]. 航空学报, 2018, 39(3): 172-183.CHEN Y G, XIONG S H, HE Q, et al. Dynamic recognition method for reliability of general aviation fleet equipment[J]. Acta Aeronautica et Astronautica Sinica, 2018, 39(3): 172-183(in Chinese). [23] 陈守煜. 可变模糊集理论与模型及其应用[M]. 大连: 大连理工大学出版社, 2009.CHEN S Y. Theory and model of variable fuzzy sets and its application[M]. Dalians: Dalian University of Technology Press, 2009(in Chinese). [24] KUMAR R K, RAO K S. A novel approach to unsupervised pattern discovery in speech using convolutional neural network[J]. Computer Speech & Language, 2022, 71: 101259. [25] PRAYUDA M F. Classification of sad emotions and depression through images using convolutional neural network (CNN)[J]. Jurnal Informatika Universitas Pamulang, 2021, 6(1): 90. doi: 10.32493/informatika.v6i1.8433 [26] 王奕惟, 莫李平, 王奕首, 等. 基于全航段QAR数据和卷积神经网络的航空发动机状态辨识[J]. 航空动力学报, 2021, 36(7): 1556-1563.WANG Y W, MO L P, WANG Y S, et al. Aero-engine status identification based on full-segment QAR data and convolutional neural network[J]. Journal of Aerospace Power, 2021, 36(7): 1556-1563(in Chinese). [27] 王帅, 黄海鸿, 韩刚, 等. 基于PCA与GA-BP神经网络的磁记忆信号定量评价[J]. 电子测量与仪器学报, 2018, 32(10): 190-196.WANG S, HUANG H H, HAN G, et al. Quantitative evaluation of magnetic memory signal based on PCA & GA-BP neural network[J]. Journal of Electronic Measurement and Instrumentation, 2018, 32(10): 190-196(in Chinese). -