Approach Pattern Automation With Quadratic Bézier Curve Trajectory Planning

Current procedures for sequencing and scheduling arriving aircraft are designed to handle 2-3 aircraft at a time and rely heavily on communication between air traffic controllers and pilots. However adherence to a First Come First Served (FCFS) order, along with the use of holding patterns or go-arounds, may lead to inefficient energy usage and unnecessary delays. Scaling the Federal Aviation Administration’s current scheduling methods to a high-density, semi-autonomous airspace, such as Advanced Air Mobility, presents possible challenges in terms of controller workload, aircraft sequencing, and cascading delays. Scheduling challenges become even more pronounced at vertiports, where aircraft can approach and land from any direction. The presented work investigates trajectory planning techniques that ensure multiple aircraft can make coordinated arrival path interceptions while adhering to assigned sequences and Scheduled Times of Arrival (STAs). Time-optimal trajectories composed of Bézier curves and Dubins path segments are developed and optimized for interception. Planned trajectories are autonomously executed by each agent in a landing sequence to meet precise STA requirements. In two experiments where 20 aircraft utilized time-optimal trajectory planning to autonomously enter an approach pattern, no aircraft violated its sequence assignment and the mean absolute deviation between arrival time and STA was under 2 seconds. The average percent error between Estimated Time of Arrival (ETA) of the Bézier trajectory and the STA was under 0.3% in both experiments.