Comparison of Evolutionary and Traditional Algorithms for Flight Routes Optimization

Comparison of Evolutionary and Traditional Algorithms for Flight Routes Optimization

Emina Salkanović

This study focuses on optimizing flight routes between airports using a Genetic Algorithm (GA). The goal is to reduce travel time and cost for both airlines and passengers. The research compares the performance of evolutionary algorithms like GA with traditional algorithms such as Dijkstra’s and A*. It also investigates how the network reacts to changes, such as adding or removing direct flights, and introduces a new feature that allows setting a fixed number of layovers. The dataset used includes over 6,800 real-world flight routes. The first version of GA was tested using only air time, while the second version used both air time and distance to reflect more realistic travel conditions. Results showed that Dijkstra’s algorithm is best when the shortest path is needed, but it is slow for large networks. A* is faster but less reliable in complex cases. GA was the most flexible and adapted well to changes, making it ideal for real-world airline planning. This research proves that GA can find efficient routes even when conditions change. It also shows that not all direct flights are necessary and that some indirect routes are already well-optimized. The findings are useful for airlines looking to reduce costs, improve schedules, and adapt to changing environments. The model developed in this study provides a strong base for future research in transportation and logistics optimization.

Vehicle Routing Problem (VRP), Evolutionary Algorithms, Traditional Algorithms, Flight Route Optimization, Genetic Algorithm (GA), Dijkstra’s Algorithm, A* Algorithm, Cost Efficiency, Network Adaptability