A genetic algorithm is presented for obtaining flight plans which minimize the fuel bum while meeting a scheduled time enroute. When valid solutions cannot be found which meet the scheduled time enroute, the method uses a secondary fitness function to find valid trajectories which minimize the time enroute. This approach includes aircraft performance constraints in addition to realistic operational constraints such as avoidance of active Special Use Airspace (SUA) and imposition of valid altitudes for direction of flight. The influence of numerical parameters and genetic operators is assessed. It was determined that simple crossover was useful for the production of step climbs. The calculated fuel burn was compared to solutions from flight operations manuals and was found to be within one percent. The genetic algorithm was able to find a global minimum corresponding to a test case involving a narrow corridor with very high tailwinds. When compared against 1394 filed flight plans across the CONUS using existing airways, the method was able to produce a reduction in fuel consumed under various altitude-for-direction-of-flight rules.
[1]
Robert Sutton,et al.
A Fuzzy Autopilot Optimized Using a Genetic Algorithm
,
1997
.
[2]
Arthur E. Bryson,et al.
Applied Optimal Control
,
1969
.
[3]
H. Erzberger,et al.
OPTIMUM HORIZONTAL GUIDANCE TECHNIQUES FOR AIRCRAFT
,
1971
.
[4]
Zbigniew Michalewicz,et al.
Genetic Algorithms + Data Structures = Evolution Programs
,
1996,
Springer Berlin Heidelberg.
[5]
Marc Schoenauer,et al.
Genetic Algorithms for Air Traffic Control System
,
1996
.
[6]
Z. Michalewicz,et al.
Genocop III: a co-evolutionary algorithm for numerical optimization problems with nonlinear constraints
,
1995,
Proceedings of 1995 IEEE International Conference on Evolutionary Computation.
[7]
J. Betts,et al.
Trajectory optimization on a parallel processor
,
1989
.