Different types of information are used to help aircraft maintain separation standards. At the lowest level, information is needed to indicate if separation standards will be violated in the near future, called a conflict. Once a conflict is detected, then conflict resolution information may be used to create a new path in which there is no conflict. Most future airspace concepts propose using computer algorithms to produce this information. Both conflict detection and resolution algorithms usually work in a pair-wise fashion: the ownership aircraft and one other aircraft. In situations where traffic density is low, this pair-wise assumption does not significantly impact operations. However, when traffic density is high, resolving one conflict may result in new near-term conflicts called secondary conflicts. These secondary conflicts may be nearer (in time) than the original conflict being addressed, so, the safety of the aircraft depends on avoiding these conflicts.
In air traffic management systems, a conflict prevention system examines the traffic and provides ranges of guidance maneuvers that avoid conflicts. This guidance takes the form of ranges of track angles, vertical speeds, or ground speeds. These ranges may be assembled into prevention bands: maneuvers that should not be taken. Unlike conflict resolution systems, which presume that the aircraft already has a conflict, conflict prevention systems show conflicts for all maneuvers. This work is a mathematical framework to analyze the correctness of algorithms that produce conflict prevention information.
Information to avoid potential conflicts involves analyzing possible maneuvers of the aircraft. There are two basic approaches to tactical airborne conflict prevention: probing and bands. In the maneuver probing approach, the pilot or controller provides an individual maneuver, which is tested to ensure the proposed trajectory is conflict- free. In the bands approach, large groups of possible maneuvers are analyzed and the pilot is presented with ranges of track angles or ground speeds, which, if taken, will result in conflict-free trajectories. Alternatively, these ranges could represent avoidance or “don’t go” zones. These ranges of guidance maneuvers are referred to as conflict-prevention information.
Conflict-prevention information may be presented directly to a pilot or controller, or it may be supplied to another automated system. In either case, since aircraft safety may be threatened by incorrect conflict prevention information, a rigorous analysis of this information is needed. To illustrate the general mathematical framework for the analysis of conflict prevention information, a two-dimensional algorithm is presented for track angle prevention bands. For completeness, a two-dimensional algorithm for ground speed preventions bands also was developed. Both algorithms have been formally verified as correct. Candidate algorithms for three-dimensional prevention bands also are being developed.
Although conflict-prevention systems have been used in several human-in-the-loop simulation experiments, and their functionality has been described in other papers, it is believed at the time of this reporting that this is the first published analysis of such information. The primary focus of this work is conflict-prevention systems for airborne operation, but there is nothing inherent in this approach that precludes use in ground-based systems.
The mathematics underlying conflict prevention systems is more subtle than expected. Instead of a trigonometric analysis that yields fourth-order polynomial equations, a vector-based approach is developed. First, the problem is divided along the dimensions of ground speed and track angle. Next, the problem is divided into near-term and intermediate-term time horizons. Then, the problem may be reduced from an N-aircraft problem to a pair-wise problem. Since the maximum number of transition points is fixed, this pair-wise division of the problem results in an algorithm that scales linearly with the number of traffic aircraft.
This work was done by Jeffrey M. Maddalon, Ricky W. Butler, and George Hagen of Langley Research Center. LAR-17874-1