Game-theoretic Resolution of Stochastic Multi-aircraft Near Midair Collisions

Abstract:
This paper presents a multi-aircraft Markov decision process congestion game to resolve multi-aircraft near midair collisions (NMACs) for small unmanned aerial vehicles (sUAVs). Two key features of this framework are: 1) it leverages the concept of strategic equilibria from game theory to define optimality in multi-aircraft near midair encounters and 2) it extends the existing NMAC metrics to stochastic formulations via the occupancy measure of a Markov decision process. This game-theoretic approach decomposes the classically centralized air traffic control objective to multiple objectives that correspond to each aircraft within the NMAC, and as result, provides an aircraft-centric notion of optimality and safety that is well-suited for distributed conflict resolutions in multi-aircraft NMACs. In addition to modeling multi-aircraft as a game, stochastic metrics that extend the deterministic notions of NMACs are explored. The safety and optimality of the Nash equilibrium multi-aircraft trajectory under a joint NMAC threat is analyzed under different NMAC thresholds and evaluation metrics. Results are simulated numerically for a representative sUAV NMAC geometry.

Did you attend Forum 81? Click the preview below to access the full paper.

The full paper is also available for purchase at the Vertical Flight Society Online Library and Store.