Optimisation of operation flight plan using genetic algorithm

• Main Author: Loh, Kai Leong.
• Other Authors: School of Mechanical and Aerospace Engineering
• Format: Final Year Project
• Language: English
• Published: 2011
• Subjects: DRNTU::Engineering::Aeronautical engineering::Air navigation
• Online Access: http://hdl.handle.net/10356/45750
• Institution: Nanyang Technological University

Commercial airline optimisation of operation flight plan requires a list of waypoints or Navaids and their corresponding flight levels in which the aircraft should fly. Although Dijkstra’s Algorithm (DA), which is currently used by Flight Focus Pte Ltd, can find the global optimal route, it is too time-consuming to perform 3-Dimensional (3D) optimisation. Current optimisation is achieved by performing horizontal (longitude and latitude) optimisation first, then using the results for vertical (flight level) optimisation. Genetic Algorithm (GA) presents itself as a potential candidate for coupled, 3D optimisation and the purpose of this project is to proof the feasibility. The scope is narrowed to horizontal optimisation where traditional GA is modified to allow such application. Modified GA represented the waypoints using genes in variable length chromosomes and performed genetic operations using multiple point crossover and mutation. Roulette wheel selection method and elitism were used in the algorithm as well. Extensive tests on the modified GA proved that it was functional and converged to an optimum that was less than 5%, average of around 1% and mode less than 0.25% with respect to global optimum values. However, results also indicated that biasness in route creation might have caused occurrences of local optimal solutions. Despite the possible drawbacks from route creation, Flight Focus Pte Ltd finds the study conclusive in proving the concept and supports future studies.