Dive Into QAR Stream Data - A Real-time Early Warning Mechanism

Date:

This is a research report on real-time flight safety early warning using QAR stream data. QAR is a quick access recorder that records various sensor data of the aircraft. The authors propose a data washing method, a data acquisition method, and a random forest classification model to predict potential over-limit events during the flight. The authors also present a real-world heavy landing case, analyzing the causes and effects of the over-limit event. The authors finally discuss the future work directions and contribution summary.

Table of Content

  • Introduction
    • Background
    • Motivation
    • Goals of Study
  • Methodology
    • Data Preprocess
    • Data Acquisition
    • How to Early Warning? A Real-world Example
    • Model
  • Conclusion
    • Future Work
    • Summary of Contribution

Preview

Download

[PDF]