簡易檢索 / 詳目顯示

回結果列表
研究生： 邱姿穎
Chiu, Tzu-Ying
論文名稱： 利用神經網絡進行不穩定進場檢測和分析之能量管理
Unstable Approach Detection and Analysis Based on Energy Management and Neural Network
指導教授： 賴盈誌
Lai, Ying-Chih
學位類別： 碩士
Master
系所名稱： 工學院 - 民航研究所
Institute of Civil Aviation
論文出版年： 2022
畢業學年度： 110
語文別： 英文
論文頁數： 64
中文關鍵詞： 能量指標 、深度學習 、監督式學習 、風險分析 、ADS-B
外文關鍵詞： ADS-B, energy metrics, supervised machine learning, risk assessment
相關次數： 點閱：120  下載：30 
分享至:
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報

    • 在全球航空安全計畫(Global Aviation Safety Plan, GASP)中，其目的一 (Goal 1) 即是希望可以降低飛航運行的風險，所以本研究將針對飛航階段中風險最高的進場及降落階段進行分析及探討。本研究的目標是希望透過飛行的歷史數據資料來偵測並分析飛機在進場及降落階段發生的不穩定進場及其原因。造成風險發生之因素除了低能見度及側風等天氣因素外，人為因素也占了很大一部份。飛行安全基金會 (Flight Safety Foundation, FSF) 也在過去的紀錄中發現，在進場及降落階段所發生的事故中，有將近七成的事故是因飛機能量的不當控管而造成的。由於飛行資料紀錄器 (Flight Data Recorder, FDR) 之記錄資訊不易取得，本研究將透過公開之ADS-B資料計算飛機之能量相關指標，探討非天氣因素對不穩定進場之影響，並使用監督式機器學習方法 — 神經網路 (Neural Network) 來檢測及預測台北松山機場之不穩定進場航班。在模型建立及參數調整後，輸入特徵將進行特徵重要性分析，且被檢測為不穩定進場之航班將做進一步探討。

    According to Goal 1 of six goals in the Global Aviation Safety Plan (GASP) [6] produced by the International Civil Aviation Organization (ICAO), reducing the risk of flight operations is important to improve aviation safety, thus this study will focus on the approach and landing phase which has the highest risk. This study aims to detect and analyze unstable approaches in Taiwan’s airports through historical flight data. In addition to weather factors such as low visibility and crosswinds, human factors also account for a large part of the risk. Flight Safety Foundation (FSF) also found that in the previous records nearly 70% of the accidents that occurred during the approach and landing phases were caused by improper control of aircraft energy [2]. Since the information of the Flight Data Recorder (FDR) is regarded as the airline’s confidential information, this study calculates the aircraft’s energy-related metrics and investigates the influence of non-weather factors on unstable approaches through publicly available sources — Automatic Dependent Surveillance-Broadcast (ADS-B) flight data as an alternative. In this thesis, a supervised machine learning method — Neural Network (NN) is used to detect and predict unstable arrival flights landing at Taipei Songshan Airport. The input features are analyzed after the model is built and tuned, and the flights detected as unstable are discussed.

    中文摘要 I Abstract II Acknowledgments III Table of Contents IV List of Tables VII List of Figures VIII I. Introduction 1 1.1 Research Background 1 International Civil Aviation Organization (ICAO) Global Aviation Safety Plan (GASP) 1 Automatic Dependent Surveillance-Broadcast (ADS-B) 3 Energy Management 4 Supervised Learning 5 1.2 Motivation and Objectives 6 1.3 Literature Review 6 Previous Studies for Flight Safety Analysis 7 Research and Analysis of Flight Operations Quality Assurance (FOQA) Data 7 Aircraft Energy Management Analysis 9 1.4 Thesis Organization 12 II. Methodology 13 2.1 Research Process 13 2.2 Data Overview 14 2.2.1 Automatic Dependent Surveillance-Broadcast Data 14 2.2.2 Weather Data Labeled by ATM Airport Performance (ATMAP) Weather Algorithm 15 2.2.3 Airport Information and Navigation Procedures—Taipei Songshan Airport (TSA) 17 2.3 Hierarchical Density-Based Spatial Clustering of Applications with Noise (HDBSCAN) 18 Density-based Clustering 18 HDBSCAN 18 2.4 Neural Network (NN) 19 The Evolution of NN 20 NN Model Components 21 2.5 Permutation Feature Importance 22 III. Data Preprocessing and Preparation 24 3.1 ADS-B Data Preprocessing 24 Data Cleaning and Conversion 24 Landing Runway Distinguishing 25 Outlier Detection 25 3.2 Weather Classification 27 3.3 Flight Parameters 28 Aircraft Historical Performance Parameters 29 Energy-related Parameters 29 Trajectory-related Parameters 30 3.4 Energy Boundary Construction 30 Upper Boundary 31 Lower Boundary 32 Ideal Energy State 32 Total Energy State Observation 32 IV. Model Training and Discussion 37 4.1 Overview of the Neural Network 37 4.1.1 Input/Output Data Description 37 4.1.2 Data Interpolation and Normalization 38 4.1.3 Hyperparameter Tuning 39 4.1.4 Model Architecture 40 4.2 Training Results 42 4.3 Unstable Approach Identification 50 4.4 Feature Importance Analysis 51 4.5 Discussion with Previous Work 54 4.5.1 Comparison Between Exceedance Analysis and Energy Management 54 4.5.2 Comparison of Approach Risk Analysis by Weather Data and Energy Management 54 V. Conclusion 57 Conclusion 57 Future Work 58 References 59 Appendix 62 Appendix A. ATMAP Weather Algorithm [16] 62 Appendix B. Recommended Elements of a Stable Approach [8] 63

    [1] "ICAO Safety Report 2021 Edition," 2021.
    [2] "ICAO Overview of Automatic Dependent Surveillance-Broadcast (ADS-B) Out."
    [3] "台灣飛安統計 2010-2019."
    [4] J. L. Ackley, T. G. Puranik, and D. Mavris, "A supervised learning approach for safety event precursor identification in commercial aviation," in AIAA Aviation 2020 Forum, 2020, p. 2880.
    [5] T. G. Puranik and D. N. Mavris, "Anomaly detection in general-aviation operations using energy metrics and flight-data records," Journal of Aerospace Information Systems, vol. 15, no. 1, pp. 22-36, 2018.
    [6] "ICAO Global Aviation Safety Plan 2020-2022 Edition," 2020.
    [7] J. Pollack and P. Ranganathan, "Aviation navigation systems security: Ads-b, GPS, iff," in Proceedings of the International Conference on Security and Management (SAM), 2018: The Steering Committee of The World Congress in Computer Science, Computer …, pp. 129-135.
    [8] "FSF ALAR Briefing Note 4.2 - Energy Management."
    [9] "IATA 2021 Safety Report," 2021.
    [10] U. S. N. T. S. Board, Flightcrew Coordination Procedures in Air Carrier Instrument Landing System Approach Accidents: Special Study. Department of Transportation, National Transportation Safety Board, 1976.
    [11] A. Scarinci, "Monitoring safety during airline operations: a systems approach," 2017.
    [12] L. Li, S. Das, R. John Hansman, R. Palacios, and A. N. Srivastava, "Analysis of flight data using clustering techniques for detecting abnormal operations," Journal of Aerospace information systems, vol. 12, no. 9, pp. 587-598, 2015.
    [13] L. Basora, J. Morio, and C. Mailhot, "A trajectory clustering framework to analyze air traffic flows," 7th SESAR Innovation Days, pp. 1-8, 2017.
    [14] S. J. Corrado, T. G. Puranik, O. P. Fischer, and D. N. Mavris, "A clustering-based quantitative analysis of the interdependent relationship between spatial and energy anomalies in ADS-B trajectory data," Transportation Research Part C: Emerging Technologies, vol. 131, p. 103331, 2021.
    [15] M. C. R. Murça, R. J. Hansman, L. Li, and P. Ren, "Flight trajectory data analytics for characterization of air traffic flows: A comparative analysis of terminal area operations between New York, Hong Kong, and Sao Paulo," Transportation Research Part C: Emerging Technologies, vol. 97, pp. 324-347, 2018.
    [16] P. Tsai and Y. Lai, "Risk Assessment of Final Approach Phase with ADS-B Trajectory Data and Weather Information using Artificial Neural Network," in 2021 IEEE International Intelligent Transportation Systems Conference (ITSC), 2021: IEEE, pp. 1245-1250.
    [17] Y. Jiao, H. Sun, C. Wang, and J. Han, "Research on unstable approach detection of civil aviation aircraft," Procedia computer science, vol. 131, pp. 525-530, 2018.
    [18] H. Lee et al., "Critical Parameter Identification for Safety Events in Commercial Aviation Using Machine Learning," Aerospace, vol. 7, no. 6, p. 73, 2020. [Online]. Available: https://www.mdpi.com/2226-4310/7/6/73.
    [19] T. G. Puranik, N. Rodriguez, and D. N. Mavris, "Towards online prediction of safety-critical landing metrics in aviation using supervised machine learning," Transportation Research Part C: Emerging Technologies, vol. 120, p. 102819, 2020/11/01/ 2020, doi: https://doi.org/10.1016/j.trc.2020.102819.
    [20] J. Merkt, "Flight Energy Management Training: Promoting Safety and Efficiency," Journal of Aviation Technology and Engineering, vol. 3, 10/31 2013, doi: 10.7771/2159-6670.1072.
    [21] "Airbus - A statistical analysis of commercial aviation accidents 1958-2016."
    [22] J. Sembiring, C. Liu, P. Koppitz, and F. Holzapfel, "Energy Management for Unstable Approach Detection," in 2018 IEEE International Conference on Aerospace Electronics and Remote Sensing Technology (ICARES), 20-21 Sept. 2018 2018, pp. 1-6, doi: 10.1109/ICARES.2018.8547140.
    [23] S. G. Kumar, S. J. Corrado, T. G. Puranik, and D. N. Mavris, "Application of Isolation Forest for Detection of Energy Anomalies in ADS-B Trajectory Data," in AIAA SCITECH 2022 Forum, 2022, p. 2441.
    [24] "EUROCONTROL Algorithm to describe weather conditions at European airports."
    [25] R. J. G. B. Campello, D. Moulavi, and J. Sander, "Density-Based Clustering Based on Hierarchical Density Estimates," in Advances in Knowledge Discovery and Data Mining, Berlin, Heidelberg, J. Pei, V. S. Tseng, L. Cao, H. Motoda, and G. Xu, Eds., 2013// 2013: Springer Berlin Heidelberg, pp. 160-172.
    [26] F. Rosenblatt, "The perceptron: a probabilistic model for information storage and organization in the brain," Psychological Review, vol. 65, no. 6, p. 386, 1958.
    [27] H. J. Kelley, "Gradient theory of optimal flight paths," Ars Journal, vol. 30, no. 10, pp. 947-954, 1960.
    [28] Z. Wang, L. Sherry, and J. F. Shortle, "Improving the Nowcast of Unstable Approaches," 2016.
    [29] R. J. d. Boer, T. Coumou, A. Hunink, and T. v. Bennekom, "The automatic identification of unstable approaches from flight data," 2014.
    [30] https://hdbscan.readthedocs.io/en/latest/index.html
    [31] https://skybrary.aero/articles/automatic-dependent-surveillance-broadcast-ads-b
    [32] https://www.faa.gov/air_traffic/technology/adsb/atc/
    [33] https://www.ibm.com/cloud/learn/neural-networks#toc-neural-net-u3voPJVU

    下載圖示 校內：立即公開
    校外：立即公開
    QR CODE