UAV conflict detection and resolution using probabilistic approach

This paper presents a real-time conflict detection and resolution algorithm based on a probabilistic method. It is assumed that aircraft are linked by real time data link like ADS-B. The approach uses a set of probability density functions that describe uncertainties from ADS-B. To calculate the probability of conflict, we use probabilistic method by using Monte Carlo simulations. Monte Carlo simulations are often thought of as too slow for real-time usage. However, in this paper, we use geometric based Monte Carlo simulations which allow for reducing computation time considerably. From the probability of conflict obtained, ‘Threat Level’ is determined between two aircraft, then, one of possible resolution maneuver options is chosen to minimize the probability of conflict. Resolution maneuver chosen as three-axis acceleration commands is added to the current acceleration and transformed into the control inputs. While the probability of conflict is in threat level zero, own aircraft is guided to the way-point. For guidance to the waypoint, proportional navigation guidance law is used. The paper finishes with several multiple-aircraft encounter simulations, illustrating the performance and properties of the proposed algorithm.