Machine Learning-Enhanced Aircraft Landing Scheduling under Uncertainties

TL;DR

This paper introduces a machine learning-based aircraft landing scheduling method that accounts for uncertainties, reducing total landing time by 17.2% and enhancing safety and automation in air traffic management.

Contribution

It presents a novel ML-enhanced scheduling approach integrating predictions into a MILP framework to handle uncertainties in aircraft arrivals.

Findings

17.2% reduction in total landing time

Effective handling of flight arrival uncertainties

Validated with real-world Atlanta ARTCC data

Abstract

This paper addresses aircraft delays, emphasizing their impact on safety and financial losses. To mitigate these issues, an innovative machine learning (ML)-enhanced landing scheduling methodology is proposed, aiming to improve automation and safety. Analyzing flight arrival delay scenarios reveals strong multimodal distributions and clusters in arrival flight time durations. A multi-stage conditional ML predictor enhances separation time prediction based on flight events. ML predictions are then integrated as safety constraints in a time-constrained traveling salesman problem formulation, solved using mixed-integer linear programming (MILP). Historical flight recordings and model predictions address uncertainties between successive flights, ensuring reliability. The proposed method is validated using real-world data from the Atlanta Air Route Traffic Control Center (ARTCC ZTL). Case studies demonstrate an average 17.2% reduction in total landing time compared to the First-Come-First-Served (FCFS) rule. Unlike FCFS, the proposed methodology considers uncertainties, instilling confidence in scheduling. The study concludes with remarks and outlines future research directions.