Monte Carlo Optimisation for Conflict Resolution in Air Traffic Control

Abstract

The safety of the flights, and in particular separation assurance, is one of the main tasks of Air Traffic Control. Conflict resolution refers to the process used by air traffic controllers to prevent loss of separation. Conflict resolution involves issuing instructions to aircraft to avoid loss of safe separation between them and, at the same time, direct them to their destinations. Conflict resolution requires decision making in the face of the considerable levels of uncertainty inherent in the motion of aircraft. We present a framework for conflict resolution which allows one to take into account such levels of uncertainty through the use of a stochastic simulator. The conflict resolution task is posed as the problem of optimizing an expected value criterion. Optimization of the expected value resolution criterion is carried out through an iterative procedure based on Markov Chain Monte Carlo. Simulation examples inspired by current air traffic control practice in terminal maneuvering areas and approach sectors illustrate the proposed conflict resolution strategy.