Robust Model Predictive Control for Aircraft Intent-Aware Collision Avoidance

TL;DR

This work tackles robust collision avoidance for multi-agent aircraft performing horizontal maneuvers by leveraging intruder intent in the form of waypoints and predicting its path via a Dubins trajectory. It develops an intent-aware scenario-tree robust MPC (RMPC) framework that handles discrete uncertainties in the intruder's angular velocity through a finite set of scenarios within a robust horizon $N_r$, while solving online in a receding-horizon manner. The main contribution is the integration of intent information into a tractable RMPC formulation, including a safety constraint with minimum separation and a Dubins-path-based intruder prediction, demonstrated via simulations showing improved safety margins and recursive feasibility under uncertainty. The approach offers a scalable, real-time collision avoidance mechanism suitable for integration into modern air traffic management, with future work aiming to compare against Dynamic Programming and reinforcement learning techniques and to extend to scenarios with delays and more complex intruder behavior.

Abstract

This paper presents the use of robust model predictive control for the design of an intent-aware collision avoidance system for multi-agent aircraft engaged in horizontal maneuvering scenarios. We assume that information from other agents is accessible in the form of waypoints or destinations. Consequently, we consider that other agents follow their optimal Dubin's path--a trajectory that connects their current state to their intended state--while accounting for potential uncertainties. We propose using scenario tree model predictive control as a robust approach that demonstrates computational efficiency. We demonstrate that the proposed method can easily integrate intent information and offer a robust scheme that handles different uncertainties. The method is illustrated through simulation results.

Figures (10)

• Figure 1: The evolution of the system represented as a scenario-tree.
• Figure 2: Geometry of two aircraft in the horizontal plane in the earth-fixed coordinate system. The black variables are the state variables and the greens are the velocities.
• Figure 3: An LSR Dubins path: An optimal connecting path of the initial state $\boldsymbol{\mathrm{s}}_0$ to the target state $\boldsymbol{\mathrm{s}}_T$.
• Figure 4: Comparing the classic MPC trajectory (black), the scenario-tree MPC trajectory (blue), and the nominal trajectory (green) without safety constraints for the ownship. The intruder Dubins path is shown in red and the green area is the destination of the ownship.
• Figure 5: The distance between the ownship and the intruder over time. The red dashed line is the minimum allowed distance. The green curve is for the nominal path.