Using Genetic Algorithms to Evaluate Aircraft Ground Holding Policy under Static Conditions

Abstract

The U.S. airline industry is facing delays due to congestion problems in the air transportation network. These delays not only lead to increased costs, but they also have safety related implications. Some of the costs related to such delays could be minimized by holding aircraft on the ground at the originating airport when congestion-related delays are expected en route or at the destination airport. Aside from reducing operating costs, the ground holding policy (GHP) also has beneficial implications on some safety concerns, since it is generally felt that it is better to hold an aircraft on the ground than in the air. The GHP is formulated as an integer programming problem and solved using heuristic techniques. However, the presence of discrete and binary integer variables increases the complexity involved when solved using traditional algorithms. Genetic algorithms (GAs) offer a powerful alternative for efficiently solving such problems. A GA is a search and optimization technique based on natural genetics and selection. This paper presents a discussion of GAs and their applicability to evaluating the GHP. Specific examples are presented to illustrate such applications. Results obtained from the test problems are consistent with expectations, indicating that GA can be used as a solution technique. It is observed that the program run time was reasonably low.